Jekyll2022-06-02T10:02:17+00:00http://ifitsmoving.com/feed.xmlIf It’s Moving, It’s BrokenA Structural Engineer's Take on the WorldSean KaneGalvanic Corrosion - Some Observations From the Field2020-01-06T00:00:00+00:002020-01-06T00:00:00+00:00http://ifitsmoving.com/2020/01/06/galvanic_corrosion<p>Part of any structural design process is selecting the material to build your structure from. This is an optimisation process, where you choose the materials based on their strength, ease of fabrication, durability, cost and other factors. Something that often gets overlooked in this process however is the potential for galvanic corrosion.</p>
<p>In mining, corrosive environments are common. Large amounts of water are used, and water is often contaminated with corrosive chemicals, and spillage of fine product holds moisture against steelwork (rather than letting it evaporate). Once mined, the product has to go somewhere, often to a port in a marine environment (plenty of salt!).</p>
<p>Therefore, an important design consideration is resistance to corrosion. Stainless steel would seem to be ideal to use in these locations. Unfortunately, stainless steel costs more than normal structural steel. Because of this, typically the following choices are made:</p>
<ul>
<li>Structural steel is used for large volume components which can be painted.</li>
<li>Stainless steel is used for small items or items that aren’t possible to paint easily, such as:
<ul>
<li>Services brackets are lightweight, thin, and often have complex profiles (low material cost combined with difficulty painting).</li>
<li>Moving parts such as pins (paint fails at the interface with stationary parts).</li>
<li>Items exposed to material flow / wear (paint would wear off).</li>
<li>Anything that gets added to the structure after painting (e.g. site run services).</li>
</ul>
</li>
</ul>
<p>And then this happens:</p>
<p><img src="/assets/images/galvanic-corrosion-1.jpg" alt="Galvanic Corrosion Due to Services Support" class="general-photo" /></p>
<h1 id="whats-the-problem">What’s the Problem?</h1>
<p>So what’s going on? Galvanic Corrosion - that’s what! But what is Galvanic Corrosion?</p>
<p>Corrosion is basically an electrochemical reaction where:</p>
<ul>
<li>One or more metals is in contact with a conductive solution called the <em>Electrolyte</em> (usually but not always water).</li>
<li>At one location, some of the metal, known as the <em>Anode</em>, undergoes chemical <em>Oxidation</em> in which an oxidating chemical (usually but not always oxygen) in the electrolyte reacts with the anode. This causes the anode to give up some of its electrons, and dissolve into the electrolyte.</li>
<li>These electrons travel to another region in the metal, known as the <em>Cathode</em>, via an electrical connection (usually direct physical contact).</li>
<li>The extra electrons allow the cathode to chemically <em>Reduce</em> molecules in the electrolyte (it transfers some of the extra electrons to the electrolyte).</li>
<li>As this is an electrical process, the circuit between the oxidising and reducing regions of the electrolyte needs to be completed to allow an electric current to flow. This occurs through further electrochemical reactions in the electrolyte.<span class="super">[<a href="#wikicorrosion">1</a>] [<a href="#opentextbc">2</a>] [<a href="#libretext">3</a>] [<a href="#efunda">4</a>]</span></li>
</ul>
<p>If both the anode and the cathode are the same material (e.g. normal corrosion of steel) different areas on the surface of the metal act as the anode or cathode. Which is which depends on local surface conditions, chemical concentrations, temperature and other factors.</p>
<p>However, where two different metals are in contact, they will have different electrical potentials. Therefore there will already be a voltage difference between them, trying to drive electrons from one to the other, even before the corrosion reaction starts. Once a corrosion reaction starts, this potential difference allows one of the metals to act entirely as the anode, and the other to act entirely as the cathode.</p>
<p>This results in the metal which is the cathode being protected against corrosion. However, for the metal which is the anode, corrosion can be very rapid because the different electrical potential between the anode and cathode acts to increase the electric current between them, increasing the rate of chemical reactions and therefore the rate of corrosion. This is known as <strong>Galvanic Corrosion</strong>, or sometimes <strong>Dissimilar Metal Corrosion</strong>.<span class="super">[<a href="#wikigalvanic">5</a>] [<a href="#acagalvanic">6</a>]</span></p>
<p>So knowing this, what happens when stainless steel and structural steel are placed in contact?</p>
<ul>
<li>The stainless steel, which is more cathodic than structural steel, becomes the cathode and attracts electrons away from the structural steel. Because it is the cathode, it is protected from further corrosion.</li>
<li>In the structural steel becomes the anode, positively charged iron ions react with the electrolyte (usually water) and turn into various iron oxide molecules. Iron oxide is not able to remain bonded to the structural steel and either dissolves into the electrolyte, or sticks to the surface as <em>rust</em>.<span class="super">[<a href="#wikirust">7</a>]</span></li>
</ul>
<p><img src="/assets/images/galvanic-corrosion-diagram.png" alt="Galvanic Corrosion Diagram" class="general-photo" /></p>
<h1 id="how-can-we-avoid-it">How Can We Avoid It?</h1>
<p>Based on the corrosion process above, there are 3x conditions required for galvanic corrosion to occur:</p>
<ol>
<li>A pair of materials that have different electrical potentials.</li>
<li>A direct electrical connection between them.</li>
<li>An indirect electrolytic connection between them.</li>
</ol>
<p>This suggests some ways of eliminating galvanic corrosion.</p>
<ol>
<li>Prevent the electrolytic connection.</li>
<li>Prevent the direct electrical connection.</li>
<li>Eliminate the different materials.</li>
</ol>
<h2 id="prevent-the-electrolytic-connection">Prevent the Electrolytic Connection</h2>
<p>The corrosion process needs an electrolyte, so eliminating the electrolyte can stop the process. This is effectively impossible to achieve in mine and port facilities, which use large quantities of water as process water and for washdown. Port facilities are near the ocean where saltwater spray from the ocean provides a constant supply of moisture.</p>
<p>However, good detailing of structural steel to minimise any areas that can trap moisture or spillage can help reduce the rate of galvanic corrosion by allowing wet areas to dry out. It won’t eliminate galvanic corrosion, but good detailing is always worthwhile, even when you’re not worried about galvanic corrosion.</p>
<p>Painting the connection between materials can also eliminate the electrolytic connection. This attempts to prevent the electrolytic connection by placing a paint coating over the interface between stainless and structural steel. However, there are a number of problems with this approach:</p>
<ol>
<li>If the connection is a welded connection, any gas bubbles, undercut, over-reinforcement or areas of poor weld profile will be locations where the paint coating is too thin, too thick (prone to cracking) or simply doesn’t stick.</li>
<li>If the connection is a bolted connection, the paint is likely to be too thick or thin over the joint line, and any movement between the parts will crack the paint.</li>
<li>If the joint is a field joint, it is unlikely that a quality paint coating can be achieved due to dust, humidity, time constraints etc.</li>
<li>One of the main reasons for using stainless steel is to avoid painting it, so the paint coating is usually only extended a nominal distance onto the stainless steel. However, some environments are wet enough that you can have electrolytic connections over even wide paint coats on the interface.</li>
</ol>
<p>A paint coating is better than nothing, but in my experience don’t expect it to stop galvanic corrosion for long. The following figures show a typical detail and failure of this same detail within ≈5 years of installation:</p>
<p><img src="/assets/images/galvanic-corrosion-paint-coating.png" alt="Painted Stainless to Structural Steel Interface" class="general-photo" /></p>
<p><img src="/assets/images/galvanic-corrosion-ineffective-paint-coat.jpg" alt="Failure of Painted Stainless to Structural Steel Interface" class="general-photo" /></p>
<h2 id="prevent-the-direct-electrical-connection">Prevent the Direct Electrical Connection</h2>
<p>The corrosion process also needs a direct electrical connection - preventing this connection will stop the corrosion. To eliminate the electrical connection:</p>
<ul>
<li>Avoid welded connections between structural and stainless steel.</li>
<li>Bolted connections need to be isolated with insulating washers, bushes & insulating tape.</li>
<li>Contact points (under floormesh / floorplate) should be isolated with isolating tapes, coatings etc.</li>
</ul>
<p class="caption"><img src="/assets/images/galvanic-corrosion-joint-isolation.png" alt="Joint Isolation" class="general-photo" />
Image from Standards Australia, <em>AS4673-2001 Cold-Formed Stainless Steel Structures</em>, Figure C5</p>
<p>If not properly isolated, bolted connections can cause significant local corrosion damage. Care also needs to be taken when installing the isolating materials. Usually these are nylon washers and thin layers of insulating tape/rubber, which can be easily damaged by over-torquing or sharp edges. The following image shows an example where the nylon washers have been cracked by over-tightening (although corrosion to date appears to be caused by paint failure, not galvanic corrosion).</p>
<p><img src="/assets/images/galvanic-corrosion-cracked-washer.jpg" alt="Cracked Isolating Washers" class="general-photo" /></p>
<h2 id="use-only-one-material">Use Only One Material</h2>
<p>The best way to prevent galvanic corrosion is to use only one material, either structural steel or stainless steel. This eliminates the electrical potential difference that causes galvanic corrosion, and avoids any concern about damaging paint coatings or forgetting to install isolating materials. If the material you choose to use is structural steel, general corrosion can still occur but it should be less aggressive than galvanic corrosion would have been.</p>
<h2 id="use-a-large-anode--a-small-cathode">Use a Large Anode & a Small Cathode</h2>
<p>While not on the list of 3 methods above, another recommendation sometimes given is to make sure that the area of the cathode is significantly smaller than the anode. The galvanic current produced during galvanic corrosion is related to the relative surface areas of the anode and the cathode. With a small cathode and a large anode, the galvanic current should be low.</p>
<p>This is used to justify the use of stainless steel fasteners (bolts, screws etc.) even without isolation. The stainless steel fastener (cathode) has a small surface area compared to the rest of the structure (the anode) so galvanic corrosion rates <em>should</em> be low. However, in my experience of inspecting structures, this is often ineffective at preventing corrosion.</p>
<p>This is because bare structural steel is almost always painted to avoid general atmospheric corrosion, regardless of the presence or absence of stainless steel fittings. Paint is never 100% effective, but early in its life it is usually very good at isolating the structural steel from any potential electrolytes. However, near the fasteners, there is a narrow ring of paint that has been damaged by the fastener itself, or the tools used to tighten them.</p>
<p><img src="/assets/images/galvanic-corrosion-small-cathode.png" alt="Galvanic Corrosion - Small Cathode" class="general-photo" /></p>
<p>The actual relative areas that need to be considered are <em>NOT</em> the entire structural steel structure compared to the stainless fasteners. Instead, they are the very small area of anode in the ring around the fastener and the (relatively) large surface area of the fastener head. This is actually a large cathode, small anode situation, and can result in very aggressive local corrosion as the following photo I took a few weeks ago shows:</p>
<p><img src="/assets/images/galvanic-corrosion-small-cathode-big-damage.jpg" alt="Galvanic Corrosion - Small Cathode, Big Damage" class="general-photo" /></p>
<p>You might be thinking to yourself though, “That’s obviously a small bolt for a services attachment, who cares?”. However, that’s only the most recent photo I had easily available. Stair treads and grating clips are common sources of galvanic corrosion that I see, and failure of a stair tread or a mesh panel could easily be fatal (see <a href="/2018/12/28/the-day-almost-a-statistic.html">this post</a>).</p>
<p>Even more concerning is the use of unprotected and un-isolated stainless steel pins in major structures. In this case, galvanic corrosion combines with movement of the joint, which works any rust products free and ensures that good metal-to-metal contact, resulting in extremely rapid corrosion. The following figures show the corrosion damage that occurred within ≈5 years of installation of stainless steel pins into a mild steel lug. These pins supported an operator’s cabin, and failure of this connection could have resulted in the death of the operator.</p>
<p><img src="/assets/images/galvanic-corrosion-cabin-pin-1.jpg" alt="Galvanic Corrosion - Operator's Cabin Pin 1" class="general-photo" /></p>
<p><img src="/assets/images/galvanic-corrosion-cabin-pin-2.jpg" alt="Galvanic Corrosion - Operator's Cabin Pin 2" class="general-photo" /></p>
<h1 id="so-when-can-i-use-stainless-steel-with-my-structural-steel">So When <em>Can</em> I Use Stainless Steel with My Structural Steel?</h1>
<p>It’s not always a bad idea to combine stainless and structural steels, if done properly. From what I’ve seen on site though, the following rules should be followed:</p>
<ul>
<li>Stainless steel doesn’t excuse bad detailing.
<ul>
<li>Always detail your structures so they don’t catch water or spillage and can drain.</li>
<li>Design structures and especially connections that can be inspected, maintained and replaced, so that when galvanic corrosion or any other damage occurs it can be found and fixed.</li>
</ul>
</li>
<li>For pins and any other items that can move, don’t combine stainless and structural steel. Any movement in the joint will damage any isolation you tried to put into place and allow corrosion to occur.
<ul>
<li>If you must use a stainless pin, use stainless lugs & clevises, and bolt them on as a unit.</li>
</ul>
</li>
<li>Avoid welding stainless steel to structural steel. This is almost always a bad idea.
<ul>
<li>Consider using sacrificial structural steel bosses / cleats welded to the structural steel, with holes drilled / tapped for stainless fasteners.</li>
<li>If welding stainless steel to structural steel can’t be avoided, use high quality welds and grind them smooth to take a good quality paint coating. Extend the paint coating as far from the joint as you can. If at all possible, do the welds and the paint coat in a workshop, not on site, for better coating quality.</li>
</ul>
</li>
<li>Ensure that all bolted joints are properly isolated.
<ul>
<li>Don’t forget the bolt shanks inside the hole, and the faying surfaces between clamped parts.</li>
<li>If at all possible, paint the stainless bolt / fitting to repair any paint damaged during installation.</li>
</ul>
</li>
<li>The “Small Cathode, Large Anode” effect may work in your favour, but don’t trust it unless you can ensure the cathode really is small compared to the real anode.</li>
<li>In my experience, a large proportion of improperly installed stainless steel items are not structural. Owners & operators of mine and port facilities need to educate <em>all</em> their tradespeople including electricians, plumbers, carpenters and other trades on what is acceptable and what isn’t when it comes to installing services on existing structures.</li>
</ul>
<h1 id="references">References</h1>
<ol>
<li><a id="wikicorrosion"><a href="https://en.wikipedia.org/wiki/Corrosion">Wikipedia - Corrosion</a></a></li>
<li><a id="opentextbc"><a href="https://opentextbc.ca/chemistry/chapter/17-6-corrosion/">Chemistry - 17.6 Corrosion</a></a></li>
<li><a id="libretext"><a href="https://chem.libretexts.org/Bookshelves/General_Chemistry/Book%3A_Chem1_(Lower)/16%3A_Electrochemistry/24.08%3A_Electrochemical_Corrosion">LibreTexts Chem1 - 24.8: Electrochemical Corrosion</a></a></li>
<li><a id="efunda"><a href="https://www.efunda.com/materials/corrosion/corrosion_basics.cfm">Efunda - Corrosion Fundamentals</a></a></li>
<li><a id="wikigalvanic"><a href="https://en.wikipedia.org/wiki/Galvanic_corrosion">Wikipedia - Galvanic Corrosion</a></a></li>
<li><a id="acagalvanic">Australian Corrosion Association (2013). Galvanic Corrosion and Dissimilar Metals. <em>Corrosion & Materials</em>, April 2013</a></li>
<li><a id="wikirust"><a href="https://en.wikipedia.org/wiki/Rust">Wikipedia - Rust</a></a></li>
</ol>Sean KanePart of any structural design process is selecting the material to build your structure from. This is an optimisation process, where you choose the materials based on their strength, ease of fabrication, durability, cost and other factors. Something that often gets overlooked in this process however is the potential for galvanic corrosion.Some Effects of Local Corrosion2019-09-08T00:00:00+00:002019-09-08T00:00:00+00:00http://ifitsmoving.com/2019/09/08/local-corrosion<p>A common issue that I see in my role as a structural engineer is corrosion (and lots of it!). After finding some corrosion, I’m often asked: “how much strength has been lost”?</p>
<p>Sometimes the answer is obvious:</p>
<p><img src="/assets/images/sometimestheanswersobvious.jpg" alt="Sometimes the Answer is Obvious" class="general-photo" /></p>
<p>But often it’s a bit trickier, especially when small pits are located on large surfaces.</p>
<p><img src="/assets/images/localcorrosionpitting.jpg" alt="Local Corrosion Pitting" class="general-photo" /></p>
<p>Since the potential ranges of pit size & location is effectively infinite, and there is a need to consider deflection, fatigue and ultimate limit states, it’s usually not possible to give an easy answer. The gold standard for determining the effects of corrosion is a Finite Element Analysis (FEA). However, this takes time and is not something that can be done rapidly on site. Therefore I thought I’d do some test models and trial some other potential methods for determining stress caused by corrosion.</p>
<h2 id="ultimate-behaviour">Ultimate Behaviour</h2>
<p>For ultimate behaviour in tension it is usually assumed that the whole section can achieve the full yield strength of the steel. Local stress concentrations, eccentric moments etc. do not matter - as the section goes plastic these stresses will redistribute (provided the section doesn’t fracture first).</p>
<p>For compression loads though, buckling may occur and significantly reduce strength. Rather than occurring uniformly along the plate, yielding concentrates in a “plastic hinge” in the buckled zone. This is heavily dependent on plate thickness.</p>
<h2 id="behaviour-in-service">Behaviour in Service</h2>
<p>When considering fatigue (I’m ignoring deflection as it doesn’t usually affect my work), any stress concentrations are very important. Fatigue life is proportional to \(\sigma^{3}\), and a 25% increase in stress halves fatigue life. Therefore the size of the corroded area, the depth of corrosion and the location (in both width & thickness) are all important.</p>
<p>To try and identify a simple approach that could be used to quickly determine the effects of corrosion, I brainstormed a number of possible ideas. The following seemed reasonable to me at first:</p>
<ol>
<li>
<p>Simply determine how much of the cross-section has been lost, and use that to determine the increase in strength.</p>
\[\sigma_{\text{corroded}} = \sigma_{\text{original}} \frac{A_{\text{original}}}{A_{\text{corroded}}}\]
<p>This has the advantages that it is very simple to calculate. However, this method ignores local effects of stress concentration and eccentricity.</p>
</li>
<li>
<p>If the area of corrosion is small enough, it may not have much effect on the overall strength of the section. However, local stresses in the corroded area could be higher. Therefore, the stress in the corroded section might be approximated by:</p>
\[\sigma_{\text{corroded}} = \sigma_{\text{original}} \frac{t_{\text{original}}}{t_{\text{corroded}}}\]
<p>Again however, this method ignores the local effects of stress concentration and eccentricity.</p>
</li>
<li>
<p>This approach could be extended to consider the local effects of section loss & eccentricity caused by the corrosion:</p>
<ol>
<li>
<p>Calculate the increase in stress caused by the loss of area, and;</p>
</li>
<li>
<p>Add the additional stress caused by any eccentricity of the section loss, approximately:</p>
</li>
</ol>
\[\sigma_{\text{corroded}} = \sigma_{\text{original}} \times \text{stress increase}\]
\[\text{stress increase} = f \left(\text{section loss}, \text{eccentricity} \right)\]
<p>Where stresses are less than yield (elastic), this can be calculated as:</p>
\[\sigma_{\text{corroded}} = \sigma_{ \text{original}} \left( \frac{t_{\text{original}}}{t_{\text{corroded}}} + \frac {6 \cdot k \cdot t_{\text{original}} \cdot \text{eccentricity}}{ t_{\text{corroded}}^2} \right)\]
<p>This approach begins to think about the effects of eccentricity on the stresses, but:</p>
<ul>
<li>
<p>If the structure is statically indeterminate, the stresses depend on the relative stiffness of each component. This is described by the variable \(k\) in the equation above. \(k\) could range between 1, if all the moments due to eccentricity are in the corroded area, and 0, if they are all in un-corroded steel. The value of \(k\) is likely to be very difficult to determine absent an FEA model.</p>
</li>
<li>
<p>Again, it ignores stress concentration effects due to the shape of the corroded area.</p>
</li>
</ul>
</li>
<li>
<p>Some sort of approximation to a simple beam could be made. However where corrosion areas are irregular it would likely be difficult to do this quickly.</p>
</li>
<li>
<p>The gold standard would be to do an FEA model. This will account for the local effects of the section loss, eccentricity of corroded area and any stress concentration effects due to the shape of the corroded area. It will also account for any global effects on the behaviour of the whole member caused by the corrosion.</p>
</li>
</ol>
<h2 id="test-models">Test Models</h2>
<p>To test out some of these approaches, I decided to put together a series of test FEA models, using the following patch of corrosion:</p>
<p><img src="/assets/images/corroded-section-sketch.png" alt="Corrosion Test Models" class="general-photo" width="75%" /></p>
<p>To look at ultimate load effects a non-linear material analysis was used, assuming the steel was elastic-perfectly plastic. The following figure shows the ultimate tension capacity test:</p>
<p><img src="/assets/images/corroded-tension-animation.gif" alt="Tension Test" class="general-photo" /></p>
<p>For stresses below ultimate, a linear-static analysis was carried out with a constant unit stress (1MPa) applied.</p>
<p><img src="/assets/images/corroded-section-linear.jpg" alt="Linear Analysis" class="general-photo" /></p>
<h2 id="so-what-did-i-learn-if-anything">So What Did I Learn (If Anything?)</h2>
<h3 id="ultimate-behaviour-1">Ultimate Behaviour</h3>
<p>Ultimate behaviour of the corroded plates was basically as expected. In tension, the capacity depended entirely on the total section loss that occurred. All configurations analysed failed at a load determined only by the remaining area of steel. The following figure shows typical load-deflection curves:</p>
<p><img src="/assets/images/corroded-tension-graph.png" alt="Corroded Sections in Tension Graph" class="general-photo" /></p>
<p>In compression there was a very wide spread of failure loads. Again, this was as expected, as compressive buckling is heavily dependent on thickness and secondary moments caused by eccentricity. In all cases, an equivalent uniform section loss gave a lower buckling resistance than the actual corroded models. Again, this was as expected because it concentrated the material closer to the centreline (where it has less ability to resist buckling) than in the corroded models (where the intact steel was distributed further away from the centreline). The following figure shows some sample load-deflection curves:</p>
<p><img src="/assets/images/corroded-compression-graph.png" alt="Corroded Sections in Compression Graph" class="general-photo" /></p>
<h3 id="behaviour-in-service-1">Behaviour in Service</h3>
<p>In contrast to ultimate behaviour, none of my approaches for determining stresses in the elastic range were accurate, even for preliminary site assessments. I was expecting them to be inaccurate, but I was astounded by just how inaccurate!</p>
<p>Using the reduction in area badly underestimates the increase in stress:</p>
<p><img src="/assets/images/corrosion-predictions-area-only.png" alt="Corrosion Predictions, Area Only" class="general-photo" /></p>
<p>Using the reduction in thickness isn’t any better, only it overestimates the increase in stresses:</p>
<p><img src="/assets/images/corrosion-predictions-thickness-only.png" alt="Corrosion Predictions, Thickness Only" class="general-photo" /></p>
<p>And trying to account for the local eccentricity in thickness overestimates things even more:</p>
<p><img src="/assets/images/corrosion-predictions-with-eccentricity.png" alt="Corrosion Predictions" class="general-photo" /></p>
<p>So far all these approaches have ignored the eccentricity in width of the corrosion. Because I didn’t want to spend more time thinking about equations, I pulled out Python’s <a href="https://scikit-learn.org"><code class="language-plaintext highlighter-rouge">scikit-learn</code></a> module. After adding some terms for eccentricity in width I performed multi-variable linear regression (<code class="language-plaintext highlighter-rouge">scikit-learn</code> may have been overkill). Adding in some terms for the the eccentricity in width resulted in significantly better estimates:</p>
<p><img src="/assets/images/corrosion-predictions-linreg.png" alt="Corrosion Predictions With Eccentricity" class="general-photo" /></p>
<p>However, peak errors were over 25% - not really accurate enough if you’re concerned about fatigue (although perhaps accurate enough for other purposes). Also, I wouldn’t want to trust the resulting equation based on the small number of scenarios I considered - I can think of a number of other variables I haven’t considered that could affect the resulting stresses.</p>
<h3 id="why-the-inaccuracies">Why The Inaccuracies?</h3>
<p>So why were my initial ideas for quickly determining the stresses wrong? And what else needs to be considered? A few things come to mind.</p>
<ul>
<li>
<p>Probably most counter-intuitive was that all the methods above assume the peak stress is in the corroded plate. The peak stress is often in the surrounding plate, and the simple methods above don’t address this at all. The most likely explanation for this is stress re-distribution between the corroded and un-corroded areas.</p>
<p><img src="/assets/images/corrosion-predictions-stress-redistribution.png" alt="Stress Re-Distribution" class="general-photo" width="75%" /></p>
<p>Note that the re-distribution of stresses around the corroded area also results in secondary stresses perpendicular to the main stress field.</p>
</li>
<li>
<p>Another observation was that the approximation using area tended to underestimate and the approximation using thickness tended to overestimate the increase in stresses. The method using area underestimates the increase in stress, because it assumes that stress is fully redistributed across the plate. The method using thickness overestimates the stress for the opposite reason - it doesn’t allow for any re-distribution of stress across the plate at all. In reality, stresses do re-distribute, being attracted to the thicker, and therefore stiffer, sections of the plate. The results are a peak stress somewhere between the 2x methods.</p>
</li>
<li>
<p>Method 3, which adds eccentricity to the thickness only approximation, is particularly inaccurate because it turns out that moments caused by the eccentricity in thickness are not taken by the corroded region. Using linear regression, the parameter \(k\) which distributes the moment between the corroded plate and the surrounding plate was found to be 0. Effectively all of the moment due to eccentricity in thickness is taken by the surrounding un-corroded steel, and the corroded section acts as a membrane carrying axial forces only.</p>
</li>
<li>
<p>There are a number of other variables that I can think of that may have affected these results. These include the shape of the corroded area such as length-to-width ratio, corner radius and irregularities. The through thickness profile may also have an effect, as corrosion is rarely as uniform as the test models above. Sharp notches at the bottom of corroded areas could cause very high stress concentrations locally.</p>
</li>
</ul>
<h2 id="final-thoughts">Final Thoughts</h2>
<p>The results above can be summarised as follows:</p>
<ul>
<li>
<p>If you’re interested in ultimate behaviour of corroded steel, the total area lost is probably the most important thing, at least in tension. In compression, buckling is also important but again, simply treating the section loss as uniform is generally good enough.</p>
</li>
<li>
<p>Below ultimate loads though, there doesn’t seem to be any simple way of determining the section loss. You’re much better off telling your clients you don’t know and getting the time to do an FEA model.</p>
<p>If you do need a quick & dirty way of figuring out the potential effects of section loss, I’d go with the simple ratio of thicknesses method listed above. At least it seems like it should be conservative (often <strong><em>very</em></strong> conservative).</p>
</li>
</ul>
<p>This was an interesting exercise to go through. It helped reinforce some things I already knew (about ultimate behaviour of corroded steel). However it also made me aware of areas where my assumptions were grossly wrong, including but not limited to:</p>
<ul>
<li>
<p>That the highest stress is always in the corroded area, and;</p>
</li>
<li>
<p>That there would be moments in the corroded area caused by any eccentricity.</p>
</li>
</ul>
<p>And it’s usually a good thing when you find out your assumptions are wrong, especially if its before something’s fallen down and killed someone.</p>
<h2 id="ps">P.S.</h2>
<p>It’s been a while since my last post. Family, work and other things have intervened to delay this one. Hopefully the next one will be quicker to get out…</p>Sean KaneA common issue that I see in my role as a structural engineer is corrosion (and lots of it!). After finding some corrosion, I’m often asked: “how much strength has been lost”?What They Didn’t Teach Me at University - Stiffness Part 22019-04-22T00:00:00+00:002019-04-22T00:00:00+00:00http://ifitsmoving.com/2019/04/22/stiffness-2<p>In my <a href="/2019/03/18/stiffness-1.html">last post</a> I began by saying that I felt that I had missed out on the importance of stiffness in structural engineering while I was at University. I also gave an example of where considering the stiffness of a structure’s supports had been a useful tool. In this post I want to continue with another example where the stiffness of the structure itself (not the supports) was important.</p>
<h1 id="a-concrete-example">A Concrete Example</h1>
<p>Concrete deterioration is a common issue in structural integrity. Material processing requires large amounts of water to process materials and keep dust down, and often contains contaminants such as salt. Both of these can aggressively attack the steel reinforcement in concrete.</p>
<p>A recent project I was part of included an assessment of a large concrete reclaim tunnel with obvious corrosion of steel reinforcement in the roof and walls. The floor did not appear to have any corrosion occurring, but did have a very unusual pattern of cracking.</p>
<p>The client was concerned that the cracking could have been due to overloading of the tunnel. To me it didn’t look like overload cracking. However, there was no other obvious cause for the cracking. Additionally, reinforcement cover measurements showed that the cover thickness was a <em>lot</em> higher than designed (3-4x the design cover). This would have left the floor significantly weaker than designed.</p>
<p>I had to analyse the tunnel to determine if the repairs to the wall / roof would be safe, so it was a simple task to check the tunnel floor as well. Interestingly the first pass of analysis showed that the floor WAS overloaded under normal stockpile loads. Now I was in an uncomfortable spot - was the model wrong, or was my initial judgement on site?</p>
<p>The model showed that the floor could be overloaded and the cover measurements suggested the floor was much weaker than designed. On the other hand, I know from experience that stockpiles are usually run well above normal height. If the floor could be overloaded with a normal stockpile I would expect significant damage (more than a few relatively minor cracks) from 20 years of constant operation at or above normal stockpile heights. Finally, the crack widths and spacing were both inconsistent with overload. So the model was probably overestimating the loads or underestimating the capacity.</p>
<p><img src="/assets/images/sketch-of-tunnel.png" alt="Sketch of Tunnel" class="general-photo" /></p>
<p>I could have spent weeks looking for every scrap of extra strength the floor could develop. Perhaps a detailed non-linear FEA model might give more capacity. I could measure the cover in more locations in case we’d accidentally measured it at the worst possible locations (we did try this but our GPR broke ☹). I could measure the stockpile density to see if it was lower than assumed. Or perhaps I could keep chasing my tail around and getting nowhere. This is where thinking about stiffness came to the rescue.</p>
<p>Concrete cracks under high loads, which reduces its stiffness. The tunnel was statically indeterminate, and any reduction in stiffness of the floor would transfer load to the walls which were only lightly loaded. One way of estimating this reduced stiffness is Branson’s equation:</p>
\[I_{ef} = \left( \frac{M_{cr}}{M^{\ast}} \right)^3 \left( I_g - I_{cr} \right) + I_{cr}\]
\[I_{ef} = \textrm{effective stiffness, } I_g = \textrm{uncracked stiffness, } I_{cr} = \textrm{cracked stiffness, }\]
\[M_{cr} = \textrm{cracking moment, } M^{\ast} = \textrm{load}\]
<p>The estimated stiffness of the floor at ultimate loads was significantly lower than the un-cracked stiffness. Even after reducing the stiffness of the walls and roof in a similar manner, a large portion of the floor load was transferred to the walls. All elements easily complied with current design standards (even with the high cover). This agreed with my judgement that the cracks in the floor were not related to overload.</p>
<p>This still left the question of why the floor had cracked in the first place. I eventually came to the conclusion that the very high level of reinforcement cover would have left the top surface of the slab very prone to shrinkage cracking. The commentary to AS3600 suggests that shrinkage cracks will likely be unacceptable if cover exceeds 100mm. In this case the cover in the floor was much larger than 100mm. The crack directions were also generally appropriate for shrinkage cracking. In consultation with others this was agreed to be the most likely cause.</p>
<h1 id="soft-on-stiffness">Soft on Stiffness</h1>
<p>So why did I miss the importance of stiffness at uni and early in my career?</p>
<p>I think the first reason may be that I went to a “practical” university. It had a high emphasis on skills that are useful early in a career and a low emphasis on the underlying theory that is often not needed until later, or not at all if you end up in project management or a non-technical role.</p>
<p>Analysis & design of many structures is often done by ignoring stiffness reductions at high load and assuming that all supports and joints are either perfect rigid or pinned. These are also easy examples to teach. After some hand analysis of simple structures and an overview of the matrix method students are shown a software analysis tool. From then on it’s so much easier to just model it, press “solve” and move on.</p>
<p>On the design side, most design is done to design codes and in Australia, at least for concrete and steel, the design codes are based on making sure your design satisfies strength and deflection requirements. The codes do not often have any (direct) requirements for stiffness. Therefore if you’re teaching a student to be an immediately “useful” engineer who can design to the codes you can ignore stiffness.</p>
<p>When you graduate and begin to practice it’s the same. You build a model in an analysis program, choosing section properties from a library of standard sections, not realising that as you choose you’re actually choosing the stiffness of your model. Then you press “solve” and get out forces and displacements for your code checks and move on without thinking about stiffness.</p>
<p>Finally, in structural integrity a LOT of design is localised around very local damage, and a lot of repairs are done in-situ, simply overplating members to restore their strength. If you are taking an existing member or connection and restoring it, you are usually not changing the stiffness too much relative to the original design, so you can often ignore it.</p>
<h1 id="whats-the-big-deal">What’s the Big Deal?</h1>
<p>In one sense this is making a mountain out of a molehill. Plenty of engineers graduate without knowing a lot about stiffness and:</p>
<ol>
<li>Do just fine without it. For simple structures, especially new structures, it may never be a big issue.</li>
<li>Get taught it by wiser older engineers in their graduate jobs (like so many other things we weren’t taught at uni…)</li>
<li>Move on to non-technical (management) positions before they would ever need to know about it.</li>
</ol>
<p>However the concept of stiffness can be very useful to have. It can be used as an additional tool in the engineer’s toolbox, to show that certain problems don’t (or do!) need further analysis. It can also prevent you from chasing your tail around in circles trying to make things stronger when in reality making them attract less load may be an easier solution.</p>
<p>Finally, most structural analysis software is based on the <a href="https://en.wikipedia.org/wiki/Direct_stiffness_method">Direct Stiffness</a> method of analysis which treats a structure as a collection of springs and solves for their combined displacements and forces. This approach assumes that the <em>stiffness</em> of the structure is fully known - the forces and displacements contain the unknowns. It seems odd that engineers should ignore the only part of the solution process they actually fully specify!</p>
<h1 id="so-what-to-do">So What To Do?</h1>
<p>Universities - at least those like mine - may need to focus a little bit more on theory (if you went to a sandstone university, maybe you dream in stiffness matrices). However even without that, simply using some more real world examples like those discussed on this blog would go a long way to reinforcing the importance / usefulness of stiffness. For example, I remember briefly learning about Branson’s equation in my uni course, with a sterile example “this is how we calculate the stiffness of this <em>imaginary</em> beam”. A real world example similar to the concrete tunnel above would have helped demonstrate the concept <em>and</em> showed a real world example of it being used, reinforcing its usefulness and importance.</p>
<p>Analysis software writers may have a role to play here as well. Many of the users of these packages couldn’t care less how they work inside or remember back to their analysis of structures courses. Perhaps their user interfaces could provide some reminders to the users that when they specify material properties or choose section dimensions they are really choosing the stiffness of their model. Simply calling the beam properties dialog box the “Beam <em>Stiffness</em> Properties” dialog may help.</p>
<p>Finally, practicing engineers should continue to do what they have always done - teach their graduates what they really need to know (in addition to what they were taught at uni)!</p>Sean KaneIn my last post I began by saying that I felt that I had missed out on the importance of stiffness in structural engineering while I was at University. I also gave an example of where considering the stiffness of a structure’s supports had been a useful tool. In this post I want to continue with another example where the stiffness of the structure itself (not the supports) was important.What They Didn’t Teach Me at University - Stiffness2019-03-18T00:00:00+00:002019-03-18T00:00:00+00:00http://ifitsmoving.com/2019/03/18/stiffness-1<p>No university course can teach you everything about engineering, and mine wasn’t an exception. One of the things that I didn’t learn about at uni (or that I missed…) was the use and importance of stiffness in structural analysis. Since uni though, and especially in the last few years I have been gaining a greater appreciation of how useful it can be.</p>
<h1 id="springs-lots-of-springs">Springs, Lots of Springs</h1>
<p>Anyone who has done high school physics probably saw the following equation:</p>
\[F=kx\]
<p>This is <a href="https://en.wikipedia.org/wiki/Hooke%27s_law">Hooke’s Law</a>, which describes the behaviour of a spring. \(F\) is the force the spring can exert, \(x\) is the displacement (the amount it is stretched/squashed) and \(k\) is the spring constant or it’s “stiffness”. In high school you probably only looked at springs that looked something like this:</p>
<p class="caption"><img src="/assets/images/springs.jpg" alt="Springs" class="general-photo" />
Image from <a href="https://en.wikipedia.org/wiki/File:Springs_009.jpg">Wikipedia</a>, License: <a href="https://creativecommons.org/licenses/by-sa/3.0/deed.en">CC BY-SA 3.0</a></p>
<p>However any <a href="https://en.wikipedia.org/wiki/Elasticity_(physics)">elastic</a> object can be described by the same equation. This includes beams, columns, floor slabs and most of the structure in a normal building. This provides a very useful concept for understanding how structures behave: they are collections of large numbers of springs, all pushed & pulled by the loads applied and in turn pushing and pulling on each other.</p>
<p>The analogy does fall apart when loads get near the ultimate capacity of a structure, where yielding, buckling, fracture and other odd things can happen. For the purposes of most basic structural engineering though, the “bunch of springs” analogy works fine.</p>
<h1 id="thinking-about-stiffness-is-useful">Thinking About Stiffness is Useful</h1>
<p>So how can considering stiffness be useful?</p>
<p>Probably the first time I considered stiffness directly would have been for determining foundation loads. Except for statically determinate structures, reactions on the foundation (or other structures) depend on the interaction of the structure and the foundation. In my experience clients in my industry don’t want to pay for geotechnical investigations to determine the actual stiffness of the foundation. So I would look at the info for their nearest borehole (usually at least 100m+ away…), dig out my textbooks, choose between “silty clay” or “clayey silt” (?!?!?!?!) and make my best (wildly inaccurate) guess of the stiffness. I would then do a <strong><em>very</em></strong> conservative (and expensive) design. The client wouldn’t like it, pay for a geotechnical investigation and give it to a different consultant to redesign the footing. Their design would be cheaper because they had the right information to start with. And I would look like an idiot. True story (not that I’m bitter about it…). So let’s skip footings.</p>
<p>When the reaction is another structure however, usual practice is to just assume restraints are either perfectly fixed or perfectly pinned and to ignore their real stiffness. This often gives good enough results but sometimes using the correct stiffness is useful, or even required to make a design work.</p>
<p>A couple of years ago now I was involved in a 3rd party review of the deconstruction of a wharf structure, which provided a number of examples of just how important considering stiffness can be. To keep it short though I’ll focus on the deconstruction of a large conveyor truss.</p>
<p>To remove the truss, the plan was to support the centre and cut the gantry in half (the full weight of the conveyor exceeded the crane capacity). At the centre (either side of the cut) the truss was restrained laterally by soft slings (which were modelled as soft spring restraints). The original truss supports were assumed to be infinitely stiff perfect pins. This resulted in the truss acting as a large cantilever under wind loads, with the majority of the wind load going to the original supports and very large longitudinal reactions:</p>
<p><img src="/assets/images/sketch-of-truss-infinite-stiffness.png" alt="Sketch of Truss - Infinite Stiffness" class="general-photo" /></p>
<p>Unfortunately, the lift procedure called for removing the northern half of the truss first (immediately after cutting it in half).This was the location of the original longitudinal restraint. The southern support (for the remaining half truss) was a tall, slender trestle, cantilevered off the berth deck level. This was clearly not designed for longitudinal loads. There was a chance the trestle would be required to withstand at least some wind load - for example if a storm picked up between the lifts, or it was left overnight. A check of its capacity showed that it was significantly overloaded even under very low wind loads.</p>
<p>Being a 3rd party reviewer, I could have just stopped there and recommended that the lift not go ahead. However, the original engineers were comfortable with the arrangement, and checking the trestle would have been a pretty obvious check for them to have done. Thinking about it, I realised that although the trestle was failing because it was a tall, slender trestle - it was a tall slender trestle! Therefore it would have very low stiffness. The standard formula for the deflection of a cantilever is:</p>
\[\Delta = \frac{Pl^3}{3EI}\]
<p>This is similar to Hooke’s law - it contains force (\(P\)) and deflection (\(\Delta\)), and the remaining terms can be re-arranged to give an equation for stiffness:</p>
\[k = \frac{3EI}{l^3}\]
<p>Using the equation above the stiffness of the trestle was found to be very low. The estimated stiffness was used for spring restraints in the model of the truss. The truss now behaved very close to a simply supported beam under wind load, with load shared approximately equally between the temporary restraint and the trestle. The longitudinal reactions were well within the trestle’s capacity.</p>
<p><img src="/assets/images/sketch-of-truss-realistic-stiffness.png" alt="Sketch of Truss - Realistic Stiffness" class="general-photo" /></p>
<h1 id="the-end---for-now">The End - For Now…</h1>
<p>Hopefully this post has been an interesting look into a real world use of stiffness in structural analysis & design. In my next post I intend to give another example, and think a little about why it took me a few years of my career before I started to understand the importance of stiffness.</p>Sean KaneNo university course can teach you everything about engineering, and mine wasn’t an exception. One of the things that I didn’t learn about at uni (or that I missed…) was the use and importance of stiffness in structural analysis. Since uni though, and especially in the last few years I have been gaining a greater appreciation of how useful it can be.Those Who Don’t Know Their (Structure’s) History are Condemned to Repeat It2019-02-24T00:00:00+00:002019-02-24T00:00:00+00:00http://ifitsmoving.com/2019/02/24/Knowledge-Management<p>Over the past month or so I have been helping a new colleague start a secondment role at a large industrial site where I have done a lot of work over the past 10 years. In the short time he has been on site there have been at least 4 incidents where damage has been identified in structures where I (as an outside consultant) know more of the history of the issues than those on site seem to (at least so far as I can tell):</p>
<ul>
<li>
<p>Cracking of a large bin. This structure had a history of cracking in a particular connection going back to (presumably) construction. This recently re-occurred. Unfortunately, this history was only documented for a specific 2 year period (out of its 10 year life).</p>
<p>The only reason the site had the limited history it did was because I had written it in a secondment a few years ago. I was able to speak to a few people who had been involved in repairs and some who had vague memories of the construction of the bin. Without knowing the history of cracking, repairs and modifications it was very difficult to determine exactly what was going on.</p>
<p>Eventually a different engineering consultant installed strain gauges and carried out an FEA analysis. They concluded the cracks would stop growing before they caused serious damage. But if I owned the bin it would still be nice to know the full history of cracking just to be sure…</p>
</li>
<li>
<p>Cracking of a materials handling machine (call it machine <strong>A</strong>). My colleague mentioned in passing that he’d looked at some cracking on a particular connection. The site team had decided to weld it up in-situ, and he had reviewed their methodology.</p>
<p>Unknown to him, this connection was identical to a connection on another machine (machine <strong>B</strong>) that I had investigated about 5 years ago due to fatigue cracks. My recommendation at the time was to install strengthening on <strong>B</strong>, <strong><em>and</em></strong> to investigate the same connection on <strong>A</strong> for cracking and possible strengthening. When I found out that cracking had occurred on <strong>A</strong> I passed the report I’d done 5 years before to him and asked him to investigate more.</p>
<p>My colleague found there was evidence of past crack repairs on <strong>A</strong> in addition to the repairs just carried out. However, no strengthening had been installed as per my recommendation 5 years ago. Nobody on site seemed to know of either the past cracking & repairs on machine <strong>A</strong>, of the existence of my report or of the recommendation to install strengthening.</p>
</li>
<li>
<p>Corrosion damage to a conveyor support. My colleague was interested to see if his current problem was a more widespread issue and found out that repairs had been done to the next support along approximately 2-3 years ago, but could not locate any information on the site’s system about the past repairs. I was able to find some photographs of the repairs in my own records (although they weren’t helpful in this situation unfortunately).</p>
</li>
<li>
<p>Corrosion damage to cladding. They did not have photographs from the last structural inspection easily accessible - again I managed to find the original copies on our system.</p>
</li>
</ul>
<p>In all these cases the site would have benefited from better tracking of their structure’s history. They would have a better idea which components of their structure were suffering from corrosion and fatigue damage, and when & why, which might allow them to carry out more effective repairs. Also, without this information, in the event of things going wrong (such as machine A suffering a significant failure due to fatigue cracking) there could have been safety hazards to personnel on site and potentially legal or insurance trouble. Finally - it would have saved the site several hours of my time and many more of my colleague’s time, keeping costs down.</p>
<p>At this particular site it’s usually not that they have lost the information entirely, just that much of it is obscurely filed away, and a number of personnel have left (such as the person I delivered my report to). It’s there to be found if people have the time to look, but inevitably when trying to run a large, busy industrial facility no-one has the time.</p>
<p>This is not to suggest that this particular site I drew the examples from is especially bad at structural integrity. In fact, of the sites I work at they are one of the more proactive when it comes to repairs and inspections. These were just a series of recent events that I happened to be aware of due to my involvement. No doubt there dozens of structural repairs over that time-frame without any issues.</p>
<h1 id="so-whats-the-solution">So What’s the Solution</h1>
<p>I’ve got a few ideas about how this site (and others) could do things better. Some of these are pretty basic, and you should bear in mind that I’m somewhat of an outsider as well - I’m a consulting engineer, I don’t work on these sites everyday so I may miss some of the things that those who work on site full time may know. However, here goes:</p>
<ul>
<li>
<p>A register or other history needs to be kept of any structural integrity issues, especially those that are likely to be repeated events (such as fatigue cracking, or damage due to repeated overloading). This should include photographs and measurements from each event and any other information that could help diagnose the cause.</p>
<p>This is one area where many sites at least in Central Queensland seem to have problems. It is a frequent occurrence that when I or my colleagues identify cracking in a structure there is also evidence of previous repairs, without any apparent knowledge on site of when the previous cracking occurred.</p>
</li>
<li>
<p>All structural inspection information needs to be stored in a logical and easily accessible location. Ideally this would be tied to the machine or structure first, rather than to the date of inspection or an arbitrary file structure on the server (often on someone’s personal hard drive space unfortunately…).</p>
</li>
<li>
<p>Any recommendations to carry out repairs or investigations need to be entered into the site’s maintenance planning / work order system immediately.</p>
<p>I have recently been reading <a href="https://gettingthingsdone.com/">Getting Things Done</a> - a personal productivity guide. One of the key steps of the GTD method is directly applicable here: don’t try and remember anything important - write it down! Or in the case of a large industrial site, don’t try and let your people remember anything, get them to put it in SAP (or whatever system you use), even just as a placeholder to make a decision later.</p>
<p>If this had been done, the site I mentioned above would hopefully have had an inspection scheduled for machine <strong>A</strong> to look for cracking.</p>
</li>
<li>
<p>After repairs have been completed a record of the repairs needs to be kept. Ideally this would include photographs, notes from supervisors, welding check sheets etc. Where structure has been modified or overplated, site measurements of the final structure should be kept and where possible as-built drawings up-dated. Having a record of what was repaired, when & how allows for better decisions to be made about how critical new damage is and reduces surprises and “uh oh, it won’t fit” moments when carrying out future repairs.</p>
<p>If this history <em>is</em> kept, this should be communicated to anyone working on structural integrity / maintenance. I only just found out (after only 10 years…) that the site I drew my examples from <em>does</em> keep at least some history of repairs in their maintenance system. Checking this before starting repairs may have helped identify potential issues like the machine fatigue cracking problem.</p>
</li>
</ul>
<p>Finally, I have to wonder why <a href="https://en.wikipedia.org/wiki/Building_information_modeling">Building Information Modelling (BIM)</a> is not used more commonly on these sites. In particular, I can see a huge amount of value in associating the information listed above (history of damage, inspections, reports etc.) directly with a component of a machine on a 3D model. Rather than dig through folders & files to find some photos that may or may not be relevant, they could simply open the 3D model, select a component or group of components and bring up the drawings, past history and inspections and any other relevant data.</p>
<p>It’s true that many of these sites were built before CAD let alone BIM. Converting drawing registers with 10,000+ drawings (100,000+ on some sites), many of which were hand-drawn, into BIM may be cost prohibitive. But for critical structures such as materials handling machines even a coarse 3D model combined with BIM could be beneficial. This could be refined over time as it is used and the machine history develops.</p>
<h2 id="not-so-fast-consultants">Not So Fast Consultants</h2>
<p>Consultants also have a role to play. Often we are pressured to complete structural integrity jobs quickly (or cheaply…), so that repairs can be done and equipment can go back into production. In the rush, information not directly related to the repairs may get left out. This may include things such as site photographs & notes documenting the original issue, investigations into the cause of damage or as-built drawings of the completed repairs.</p>
<p>Maybe we should be willing to push back against our clients a little and suggest that perhaps they really should get an as-built drawing or a proper investigation done this time. If the client doesn’t want / isn’t pushing for this information it is probably unlikely that they’ll pay for it, and doing too much work for free sends consultants out of business - so it’s probably unlikely that there’ll be a radical change in how things work. But maybe we can try a little harder to push our clients to better practices. And where it doesn’t really cost us anything we could just share any information collected with our clients - how hard is it to put our photographs up on dropbox / google drive and email a link?</p>
<h1 id="the-end-of-history">The End of History</h1>
<p>Keeping the history of your structures is an important part of managing their structural integrity. Without it, you may be condemned to repeat it! So if you’re in charge of structural integrity, start a register today with the corrosion you found last week, or put a reminder in SAP to follow up that report you don’t have time to read. If you’re a consultant, think about sending your last inspection photos to your client.</p>
<p>Or don’t - it’ll mean more work for me when things break, so I don’t mind either way 😁</p>Sean KaneOver the past month or so I have been helping a new colleague start a secondment role at a large industrial site where I have done a lot of work over the past 10 years. In the short time he has been on site there have been at least 4 incidents where damage has been identified in structures where I (as an outside consultant) know more of the history of the issues than those on site seem to (at least so far as I can tell):The Day I Almost Became a Statistic2018-12-28T00:00:00+00:002018-12-28T00:00:00+00:00http://ifitsmoving.com/2018/12/28/the-day-almost-a-statistic<p>About 18 months ago now I came as close as I ever want to be to being an industrial accident statistic. A fairly simple inspection of a small walkway almost ended up in disaster. This post shares some of my thoughts about it, in the hopes that it can help someone else avoid a similar incident.</p>
<h1 id="what-happened">What Happened</h1>
<p>I was inspecting a conveyor structure with one of our graduate engineers, taking measurements for a repair strategy we were developing. While walking back along the conveyor towards the cribroom (we had finished for the day) we noticed that there was a heavily corroded access platform on one of the conveyor trestles. Both of us were concerned that it might be unsafe, so I decided that we needed to take a closer look at it.</p>
<p>At the base of the access ladder we looked up and our concerns were not eased - it looked even more corroded than from a distance. Turning to the graduate I jokingly asked him to go up and have a look for me. I can’t remember exactly what he said to me but it was clear he wasn’t going to climb the ladder and that he thought I was an idiot if I was going to.</p>
<p>However, after doing inspections of corroded steel for 8 years I was a little less afraid of rust than him. So I climbed the ladder to the underside of the platform for a closer look. The structure was heavily corroded but it looked like it would hold my weight. The floor mesh load bars looked like they had some surface corrosion but were approximately the right thickness, suggesting section loss wasn’t heavy. Much of what had appeared to be corrosion from below appeared to be coal stuck between the grating and dust stains on the mesh. I called out to our graduate that I was going to climb up the last few rungs onto the platform. He yelled back asking me to wait until he’d stepped out from under the platform so that I didn’t knock coal all over him. I think by now we can all see where this is going…</p>
<p>At the last second I thought to myself that it might be a good idea to give the platform a good kick, to see if anything happened, or at the least to clear out some of the heavy spillage that was on top of it. The first kick didn’t do anything except clear some coal through the mesh. I decided to give it a second kick to be certain, made sure I had a solid hold of the ladder stile, pulled up my leg and kick downwards - and almost fell straight through the panel of mesh.</p>
<p>Hanging onto the ladder with one hand and one leg trying to stop myself falling, I watched a shower of coal fall down where our graduate had been standing seconds before and saw the panel of mesh tilting on its corners through the floor trimmers. Grabbing with one hand I managed to snag the mesh before it fell. I was now hanging on for dear life, with one hand on the ladder, the other holding the mesh, one leg on the ladder and one leg dangling below the floor.</p>
<p><img src="/assets/images/should-that-hole-be-there.jpg" alt="Should that hole be there?" class="general-photo" width="75%" /></p>
<p>Our graduate looked up at me with a look of “what the heck do we do now” mixed with a large doses of “I told you so” on his face. After a few seconds we had both figured out that we were OK, apart from me still trying to hold onto the mesh. Luckily, about 2 minutes after the incident, someone drove up with a length of rope in his ute. We lowered the mesh to the ground, and the investigations began.</p>
<p><img src="/assets/images/let-the-investigations-begin.jpg" alt="My Attempted Assassin" class="general-photo" width="75%" /></p>
<h1 id="why-did-it-happen">Why Did it Happen</h1>
<p>Enough of the what though, now for the “Why”? Why did this happen, what have I learned and what can others learn from it?</p>
<h2 id="dont-trust-the-rust">Don’t Trust the Rust!</h2>
<p>The direct cause of the incident was that the floor mesh had corroded through and I didn’t notice it. In this case as best as I could remember after the incident I remember thinking that the load bars looked like they were corroded, but that they seemed to be close to full thickness with only surface rust on most bars. Only a few had seemed heavily corroded from below. After the incident though was obvious that the majority of load bars had completely corroded through. So how did I not notice this despite having 8 years experience inspecting structures? I identified a couple of factors:</p>
<p>Looking at the bars, some bars looked like they were a constant thickness, but on closer inspection they had a colour variation across their thickness. The change in colour indicated that some of load bar thickness was actually rust. The following image shows a bar where 1/2 the thickness is rust (and hence has no strength). However the overall thickness was approximately the same as the original load bar thickness. With coal dust on the mesh (before it collapsed) and looking up from the ladder into the afternoon sunlight the change in colour would not have been as visible, making it hard to identify the section loss.</p>
<p><img src="/assets/images/floormesh-detail-1.jpg" alt="Floor mesh detail" class="general-photo" width="75%" /></p>
<p>The second reason I missed the corrosion is that much of the damage appears to have occurred above the supporting floor trimmer. In this location coal buildup between the bars and the top flange of the floor trimmer would have prevented it from being visible.</p>
<p><img src="/assets/images/floormesh-detail-2.jpg" alt="Floor mesh detail 2" class="general-photo" width="75%" /></p>
<p>The following sketch shows how this could have been difficult to see from below:</p>
<p><img src="/assets/images/floormesh-detail-3.png" alt="Floor mesh detail 3" class="general-photo" width="75%" /></p>
<p>Rust is deceptive - don’t trust it! In this case I was willing to bet my life that the rust was only surface rust - and it was effectively full section loss. On other occasions I would have bet large amounts of money that section loss was extreme, yet when blasted effectively no section loss had occurred. Removal of spillage and mechanical cleaning or grit blasting should be carried out if at all possible before inspection.</p>
<h2 id="danger-tape-and-barricades-arent-good-long-term-controls">Danger-Tape and Barricades Aren’t (Good) Long Term Controls</h2>
<p>Other than corrosion, the other main factor in my near miss was the failure of barricading to prevent me accessing the platform.</p>
<p>About 5 minutes after I had climbed back down to ground level the graduate turned to me and said “Oh, by the way, I think I asked for that platform to be barricaded off about 12 months ago” (great timing…). Looking at the ladder in more detail we then identified that each style had a small amount of torn off danger-tape wrapped around them:</p>
<p><img src="/assets/images/old-danger-tape.jpg" alt="Old Danger Tape" class="general-photo" width="75%" /></p>
<p>And the handrails for the walkway providing access to the base of the ladder had paint scraped off where a scaffold pole would have been placed across as a barricade:</p>
<p><img src="/assets/images/scaff-clamp-marks.jpg" alt="Scaff Clamp Marks in Paint" class="general-photo" width="75%" /></p>
<p>As a short term measure, the use of danger-tape and scaffold to barricade a hazard is generally a safe & reliable way of reducing risk. The following image shows the “Safety Hierarchy of Controls” (<a href="https://www.safeworkaustralia.gov.au/risk">Safe Work Australia</a>), a common means of assessing safety controls. Barricading off the ladder would have counted as <em>isolating</em> the hazard from people. In many cases this is likely the most effective short term control.</p>
<p><img src="/assets/images/hierarchy-of-risk-control.jpg" alt="Hierarchy of Risk Controls" class="general-photo" width="75%" /></p>
<p>However, long term, is simply barricading off the structure an effective control? My near-miss shows that this may not always be the case.</p>
<p>What I eventually concluded is that something like the following sequence of events probably occurred:</p>
<ol>
<li>People forgot that the platform was barricaded off (and why).</li>
<li>High winds blew the danger-tape and associated information tags off the scaffold pole & ladder.</li>
<li>The site scaffolding contractor carried out one of their regular clean-ups of old scaffolding. With no danger tape and no other record of why the platform was barricaded, the scaffolders assumed the barricade was no longer required and removed it.</li>
<li>The hazardous platform was now exposed for unwary engineers to access…</li>
</ol>
<p>In the long term, danger tape is fragile and easily damaged, and scaffold is exposed to the general site hazard of being scavenged for other projects elsewhere. Systems to maintain / replace tape & barricades are <em>administrative</em> systems. Looking at the hierarchy of controls above, administrative controls are the 2nd weakest / least reliable form of controls.</p>
<p>Additionally, long term, even if the scaffold poles & danger tape prevent personnel from accessing a hazard, the hazard can become forgotten. This can then re-expose personnel to the hazard when the decision is made to rectify it.</p>
<p>For example, I have recently been involved in an assessment of a moth-balled facility. Large areas of the plant were barricaded off with scaffold & danger tape, and marked as unsafe. However, since it had been approximately 5 years since the plant was moth-balled, no-one knew why any of these areas were barricaded off. I was required to inspect these areas to determine what the hazards were and why areas were unsafe for access. In some areas the issues were clear. However in others I couldn’t identify any structural issues - but is this because there weren’t any issues or because I couldn’t see them? Were there mechanical / electrical hazards? How much danger did I place myself in? How much danger is still present? Who knows? It’s probably less than before I did the inspection but that’s about all I can say.</p>
<p>Danger tape and scaffold is probably going to continue to be used to isolate hazards - in many cases there probably aren’t any other solutions. However, this case identifies that there needs to be a robust <em>administrative</em> system in place to ensure that:</p>
<ol>
<li>Danger tape is maintained in good condition with clear and accurate info tags.</li>
<li>Scaffold is regularly checked to ensure that it is in place and in good condition.</li>
<li>The reason for barricading off the structure needs to be recorded. Ideally something like a site-wide register of long term barricades would be maintained, with someone responsible to ensure its accuracy, the condition of the controls and that action is being taken to remove the hazards.</li>
</ol>
<h2 id="structures-that-get-used-get-loved">Structures That Get Used Get Loved</h2>
<p>Structures that get used get loved, and those that don’t deteriorate. The platform I almost fell from had once been used to access some electrical or mechanical items. However, these had been removed many years before. Once these were removed, there was no reason for personnel to access the platform. Un-used structures can deteriorate more rapidly for a number of reasons:</p>
<ul>
<li>There is no-one to carry out minor maintenance (“While I’m repairing this thing I’ll just tighten that loose bolt”), and reduced opportunities for major maintenance. Structural maintenance is often carried out at the same time as major mechanical / electrical shutdowns. If there are no electrical or mechanical components to maintain, then opportunities for structural maintenance diminish.</li>
<li>
<p>There is no production incentive to maintain the structure. While equipment is present, personnel have to be able to access it to maintain it and keep the facility producing whatever it produces. Once the equipment is gone, the condition of the redundant structure has no direct impact on production, and maintenance can be delayed without direct cost impact.</p>
<p>TO BE CLEAR: I am not implying that in the case of my platform that its unsafe condition was a conscious decision of anyone. I personally know many of the people who likely made the maintenance decisions about the platform, and know they would not have consciously placed anyone in danger. On the other hand, I also know the time & budget pressures that many mining & industrial facilities work under. The production incentive must at least be a factor in decision making - even if only sub-consciously.</p>
</li>
<li>There is no-one to wash down spillage. In my experience it’s difficult enough to get operators to wash down walkways that they use, let alone walkways they don’t. Spillage traps moisture and contains contaminants that cause accelerated corrosion, and prevents easy inspection for damage. Section loss can be very aggressive in this environment - I have personally measured corrosion rates of >0.5mm / year. A typical floor mesh load bar is only 3mm thick and un-maintained could have a life of less than 3 years.</li>
<li>There is no-one to identify the deterioration. When personnel use a structure they have an incentive to notice and report damage - their safety depends on it. Additionally, although most sites have a structural inspection regime, time & access prevent full inspection of almost every structure. Operators & maintenance personnel are extra eyes, identifying issues that inspectors do not have the time or access to see.</li>
</ul>
<p>To address the risk of increased deterioration, redundant / unused structures should either:</p>
<ul>
<li>Continue to be actively maintained, with regular wash-down, maintenance and inspection.</li>
<li>Be removed, either at the same time that they cease to serve a function, or at some time before they can deteriorate to an un-acceptable state.</li>
</ul>
<p>If the choice is made to maintain a redundant structure, care needs to be taken to ensure that it really is maintained as well as any other structure on site. In particular, it may be difficult to achieve adequate wash-down and minor maintenance - will the night-shift operator get dirty and cold washing down the useless platform to nowhere that he never uses, or will he spend 15 minutes extra staying warm and cleaning his cabin windows? On the other hand, removal is <em>elimination</em> of the hazard - the most effective way of addressing the risk posed by deterioration of the structure.</p>
<h1 id="final-thoughts">Final Thoughts</h1>
<p>Personally, I’m surprised at how little this incident has affected me. Until writing this post I haven’t really spent any time thinking about it. There was no time to be afraid or do anything but react when the mesh fell out underneath me. I was buzzed with adrenalin in the aftermath, but by the time I got home in the evening I had very little emotional response.</p>
<p>Maybe it confirms the stereotype of the cold, mathematical, un-emotional engineer. Thinking about it a little more though it’s probably just a human thing. It didn’t kill me and there were no other negative consequences so why should I spend time and energy worrying about it? We all have moments where we escape injury or death and don’t think twice. How many times have you nearly been in a car accident and thought to yourself “pheeewww that was close” and then forgotten it?</p>
<p>It also hasn’t put me off working in mining or heavy industry. While every near miss, injury or fatality is important, I’m about 6 times more likely to get injured on the way to or from work than on site. There were <a href="https://www.safeworkaustralia.gov.au/statistics-and-research/statistics/fatalities/fatality-statistics">188 Industrial Fatalities in 2017</a> compared to <a href="https://bitre.gov.au/statistics/safety/">1224 road accident deaths</a>. This simple comparison is probably an abuse of the statistics, but it still appears likely that I face significantly more risk commuting than working (especially because I spend much of my time in the office anyway).</p>
<p>I’m glad that I was the one who saw the platform and chose to inspect it - I was at least worried about it. A night-shift electrician told to fix a light bulb may have simply climbed the ladder and stepped out into nothing without any warning. I’m also glad that I went up the ladder, not our graduate - it’s a cliche that you shouldn’t ask someone to do something that you’re not willing to do yourself but it’s a cliche for a reason.</p>
<p>Am I a more conservative engineer because of the near miss? I’m not sure. I think I’m a little less trusting of floor mesh and more willing to ask people to barricade things off. But overall though I’m not sure - I still have to think about it a bit more.</p>
<p>Finally, for those of you who are responsible for structures or safety on site, remember:</p>
<ul>
<li>Don’t trust the rust!</li>
<li>Danger tape and scaffolding are useful in the short term, but as long term controls for hazards they are actually reliant on administrative (and therefore unreliable) support.</li>
<li>Structures that don’t get used deteriorate - if you can, you should probably remove them rather than leave them to become a hazard.</li>
</ul>Sean KaneAbout 18 months ago now I came as close as I ever want to be to being an industrial accident statistic. A fairly simple inspection of a small walkway almost ended up in disaster. This post shares some of my thoughts about it, in the hopes that it can help someone else avoid a similar incident.Monorails and Flexural-Torsional Buckling2018-12-09T00:00:00+00:002018-12-09T00:00:00+00:00http://ifitsmoving.com/2018/12/09/monorails<p>Shortly after writing my previous post on <a href="/2018/09/29/what-is-structural-integrity.html">Structural Integrity</a>, I came across a great real world illustration of why engineers who specialise in structural integrity are valuable. It was also interesting because it made me realise that something I had thought was impossible actually is both possible and does occur in real structures.</p>
<h1 id="lets-talk-about-monorails">Let’s Talk About Monorails</h1>
<p>Monorails are everywhere in industrial structures - a typical processing plant that I often work in would have 40-50 scattered around the site. These are necessary because pumps, motors, pipework, conveyor pulleys etc. are too heavy to move by hand and the congested nature of these structures prevents access by crane. The monorails themselves are usually very simple structures consisting of a standard “I” section, with bolted supports in the top flange to allow a free bottom surface for a standard beam trolley to run.</p>
<p><img src="/assets/images/Monorails.jpg" alt="Monorails" class="general-photo" /></p>
<h1 id="where-the-engineer-comes-in">Where The Engineer Comes In</h1>
<p>I was at the particular site in question for a different reason, but while walking past a particular monorail the site Structural Integrity Engineer pointed to a monorail and started telling me a story.</p>
<p>While walking through the plant a few months earlier he had come across a couple of workers who had just finished using the monorail. Talking to them in passing, they had mentioned that while they were using the monorail they had noticed that it was deflecting sideways, and not downwards (as they expected).</p>
<p>Monorails are usually used to lift the same object over and over again. Presumably this had been happening every time it had been used since the early 1970s, without anyone ever thinking to wonder about it. However, the site engineer was able to recognise that what the workers were seeing was the onset of <em>Flexural-Torsional</em> buckling of the monorail beam.</p>
<h1 id="buckling">Buckling</h1>
<p>So what is buckling, and why do engineers worry about it? Buckling happens when structures are loaded in compression. They can either get shorter by shrinking a little, or by deflecting sideways (their overall length stays the same but the distance between load and support shortens). Below a load known as the critical buckling load (\(P_{crit}\)) the structure shrinks. Above this load the deflection mode governs.</p>
<p>Unfortunately, the deflection results in the structure being offset from the original compression force. Newton’s law about every force needing an equal and opposite reaction therefore results in a bending moment. The bending moment results in greater deflection, which results in more bending, which results in greater deflection … until your structure falls over.</p>
<p><img src="/assets/images/Buckling-sketch.png" alt="Buckling Sketch" class="general-photo" /></p>
<p>Buckling failure is dangerous because it can happen suddenly, without warning and without giving time to reduce load or otherwise prevent failure. The list of structures that have failed due to buckling is a long and not-so-illustrious one. For example in Australia the collapse of the <a href="https://en.wikipedia.org/wiki/West_Gate_Bridge">West Gate Bridge</a> during its construction which killed 35 workers was largely the result of buckling that occurred due to inadequate design.</p>
<p>Flexural-torsional buckling occurs because bending places the top flange of a beam into compression. Just like a column in compression, the compression flange of a beam can buckle. The web & tension flange of the beam force the compression flange to buckle sideways, which results in the whole section twisting (hence “torsional” buckling).</p>
<p><img src="/assets/images/Flexural-torsional-buckling-sketch.png" alt="Flexural Torsional Buckling" class="general-photo" /></p>
<p>This sideways movement of the top flange was what the workers using the monorail had observed. The monorail was effectively at the point of failure at the normal loads that these workers had been placing on it.</p>
<p>It appears that it was simple luck that had prevented workers loading the monorail past the critical buckling load. Without the site engineer present to recognise the issue, it would have continued to have been used, all the while presenting a risk of collapse and injury to workers.</p>
<h1 id="so-what-did-i-learn">So What Did I Learn?</h1>
<p>While an interesting observation in why engineers are necessary, what made this really interesting for me is that I hadn’t thought flexural-torsional buckling was possible for a monorail.</p>
<p>If you have a look at the sketch above, when the load is on the top flange of the beam, the resulting torsional moment \(M=P\Delta\) is de-stabilising and will twist the beam in the same direction that the buckled top flange does, resulting in failure. However, if the load is below the centre of the section the torsion will tend to counteract the effect of the buckling top flange and will try and straighten the section back out. Monorails are almost always loaded on the bottom flange (as was the monorail in question above).</p>
<p>I had naively assumed that this stabilising effect was at least as strong as the tendency of the top flange to buckle, therefore preventing flexural-torsional buckling. However, the story from the site engineer suggested that I was wrong.</p>
<p>Being interested in whether I was wrong or not I built a test model in an FEA package, and lo and behold the monorail beam fails with a very distinctive flexural-torsional buckling of the top flange (the sudden jump to the right approximately ½ way through the gif):</p>
<p><img src="/assets/images/Flexural-Torsional-Buckling.gif" alt="Flexural-Torsional Buckling GIF" class="general-photo" /></p>
<p>The following figure shows the force-displacement curve for various load heights and a case where buckling has been prevented (forcing failure by yielding of the cross section):</p>
<p><img src="/assets/images/monorail-force-displacement.png" alt="Monorail Force-Displacement" class="general-photo" /></p>
<p>As expected when the load is applied to the top flange there is a reasonably significant decrease in capacity (13% lower than load applied at the centre). When the load is applied at the bottom flange there is only a 4% increase in capacity. This is much smaller than I had expected (I was expected no or very little buckling with close to section capacity). While the exact numbers would vary considerably depending on the load and structure being analysed, at least in this case there is effectively no benefit to the structure from the load being below the centroid of the section.</p>
<p>For those who are not engineers, the sharp nature of the force-displacement curves at the critical buckling load, followed by the sudden drop-off after buckling indicates the dangerous nature of buckling failures. Typically structural loads cannot be reduced in time to prevent the structure following the curve downwards to collapse. In comparison, the section failure mode has a much flatter curve, with significant deflection before final ultimate collapse (which would occur well off the edge of the chart). The high deflection prior to ultimate failure may allow the structure to develop other load paths. It may also give the occupants some warning that failure is about to occur, possibly allowing them to reduce load before ultimate failure occurs.</p>
<p>I expect that for many engineers the fact that flexural-torsional buckling can occur even when loads are below the centroid will not make any difference to their day-to-day practice. AS4100-1998 (the current Australian Standard for steel design) does not provide easy methods to determine how much extra capacity a load below the section centroid provides so typical practice would be to conservatively assume the load is actually at the centroid. However, realising that loads below the centroid do not prevent buckling may be important for those who have to assess existing structures or deal with them on site where signs of buckling may occur.</p>
<h1 id="summary">Summary</h1>
<p>The example of a buckling monorail is an interesting real-world example of why engineers are useful, not just in design but also in the day-to-day operations of structures. It’s also a great lesson in not taking things for granted. I don’t know why I had assumed that flexural-torsional buckling is not possible when the load is below the centroid. However I am grateful that I learnt that it <em>is</em> possible through someone else’s near miss, not my own. Hopefully I’ll always be so lucky!</p>Sean KaneShortly after writing my previous post on Structural Integrity, I came across a great real world illustration of why engineers who specialise in structural integrity are valuable. It was also interesting because it made me realise that something I had thought was impossible actually is both possible and does occur in real structures.What is Structural Integrity Anyway?2018-09-29T00:00:00+00:002018-09-29T00:00:00+00:00http://ifitsmoving.com/2018/09/29/what-is-structural-integrity<p>The company I work for specialises in Structural Integrity, which is a niche area of Structural Engineering. At university I never even heard the phrase “structural integrity” let alone learnt anything about it. A google search shows a few references to structural integrity in a mechanical or aeronautical context, a lot of engineering companies claiming to provide structural integrity services and not much else. A search on <a href="https://www.amazon.com.au/s/ref=nb_sb_noss?url=search-alias%3Daps&field-keywords=structural+integrity">Amazon</a> shows that there isn’t even an introductory textbook.</p>
<p>Given this lack of information, I thought that this might be a good place to provide a brief introduction to what structural integrity is, and why specialists in structural integrity are required.</p>
<h2 id="what-is-structural-integrity">What is Structural Integrity</h2>
<p>Where google and academic sources let us down, as usual <a href="https://en.wikipedia.org/wiki/Structural_integrity_and_failure">Wikipedia</a> provides a reasonable definition of structural integrity:</p>
<blockquote>
<p><em>“Structural integrity is the ability of an item—either a structural component or a structure consisting of many components—to hold together under a load, including its own weight, without breaking or deforming excessively. It assures that the construction will perform its designed function during reasonable use, for as long as its intended life span. Items are constructed with structural integrity to prevent catastrophic failure”</em></p>
</blockquote>
<p>I think that this is a very helpful definition. It is simple, and easy to understand. Importantly, it also identifies that structural integrity is not just concerned with safe operation of structures in normal conditions, but also that they should be able to fail safely.</p>
<p>For example, is this good or bad structural integrity (again from <a href="https://en.wikipedia.org/wiki/Boeing_B-52_Stratofortress#Overview">Wikipedia</a>)?:</p>
<p><span style="display:block;text-align:center">
<img src="/assets/images/Boeing_B-52_with_no_vertical_stabilizer.jpg" alt="Is this good or bad structural integrity?" />
</span></p>
<p>This aircraft has obviously suffered a significant loss of structural integrity. Without the vertical stabiliser it is likely unable to carry out any of its original design aims (fly to a target, manoeuvre, drop bombs etc.). The cause of the failure was fatigue caused by turbulence (hardly unexpected for an aircraft). Therefore, in one sense this is bad structural integrity: the aircraft was not able to withstand reasonable design loads for its full intended design lift without breaking or deforming excessively.</p>
<p>However, importantly (especially for the crew) it is still able to maintain the bare minimum functions of an aircraft: remaining in the air and stable. According to Wikipedia the aircraft landed safely and even returned to service after repairs! Even if it could not be landed safely, simply maintaining stability for a few minutes likely would have bought time for the crew to safely eject. So the aircraft had at least a minimal amount of structural integrity.</p>
<p>Most structural engineers will never design an aircraft fuselage, but this distinction between normal operations (where no damage is expected) and extreme load events (where prevention of catastrophic failure is the goal) is still important.</p>
<p>This distinction is also useful to know for non-engineers who are often those responsible for day to day operation of structures. It is often easy to ensure that a structure is safe for normal loads. Not too many people start removing columns out of buildings without engineering advice, as everyone has a concept of gravity and things falling down from their everyday experience. However, it is easy to compromise structural integrity against catastrophic events without realising it. For example, removing bracing from structures usually doesn’t result in immediate catastrophic failure. Structures with significant damage to bracing may remain “safe” until an event such as an earthquake or cyclone exposes the weakness of the structure. Just because a structure doesn’t fall down when it is modified doesn’t mean that it hasn’t had significant reductions in its structural integrity and safety.</p>
<h2 id="why-specialise-in-structural-integrity">Why Specialise in Structural Integrity?</h2>
<p>Ensuring structures have structural integrity sounds like something that all structural engineers should be doing. All design codes require design to prevent damage in normal operation, and most include specific clauses that ensure a minimum level of robustness under extreme conditions. Even where codes don’t cover structural integrity a professional engineer should ensure they address these issues as part of their obligations to the profession and society. So why are there engineers who specialise in structural integrity?</p>
<p>The reason is that structures have a life of their own. At a recent conference I heard the quote:</p>
<blockquote>
<p><em>“Concrete doesn’t know about the code”</em></p>
<p>- Attributed to T. Paulay</p>
</blockquote>
<p>And this applies equally to all structures (not just concrete). Structures will only ever have perfect structural integrity in the mind of the designer.</p>
<p>Over their lifetime structures:</p>
<ul>
<li>May have design flaws, whether minor calculation errors, fundamental mis-understandings of structural behaviour or errors in documentation.</li>
<li>May not have been constructed in accordance with the design information specified by the engineer.</li>
<li>Deteriorate due to corrosion, fatigue, wear, rot, abrasion and general aging of all kinds.</li>
<li>Are modified, often by those without an engineering background and without consideration of the principles of structural integrity (or even basic structural mechanics).</li>
<li>Are damaged due to overload, impact, earthquakes, cyclones, explosions and other extreme events.</li>
</ul>
<p>Also, structures may have been designed before the principles of structural integrity, structural mechanics or material behaviour were well understood, and extreme events may reveal fundamental issues with whole classes of structures. Design to prevent catastrophic collapse as fundamental structural engineering practice is largely the result of a number of singnificant collapses in the 1960s-1980s, and extreme earthquake events often expose flaws in whole categories of structures. Even if a structure complied with the design codes at the time it was constructed, there is a reasonable chance that changes to codes now mean that it does not comply.</p>
<p>This is where engineers who specialise in Structural Integrity find their niche. We:</p>
<ul>
<li>Carry out inspections to identify damage and deterioration.</li>
<li>Assess structures for compliance with design codes and / or for robustness against catastrophic collapse.</li>
<li>Assist in assessing the risks that issues identified pose to the owners, operators and occupiers of structures (and the public).</li>
<li>Design repairs for issues identified.</li>
<li>Help develop systems to ensure that inspections and maintenance are carried out and that modifications do not affect structural integrity.</li>
</ul>
<p>In a sense many (maybe even most) structural engineers do a little of each of these throughout their career. However what makes an engineer a specialist in structural integrity is the focus on these issues, rather than on design, construction or other areas of engineering. Due to this focus, a structural integrity specialist will have greater experience at identifying issues in existing structures and how to address them given the constraints posed by existing structures. For example:</p>
<ul>
<li>Many structural deterioration mechanisms are hard to identify for those without experience looking for them. It takes repeated regular experience to understand what you are looking at and (importantly) where to look.</li>
<li>Quickly and efficiently determining if identified defects are important for the integrity of structures requires understanding where and how deterioration will occur in a way that design & construction engineers may not have.</li>
<li>How structures are used in reality by their owners, occupiers and operators often varies greatly from what the original designer may have intended, especially as time passes. A structural integrity specialist with expertise in a particular industry may sometimes understand how the structure is used better than the original designer.</li>
<li>The constraints posed by the geometry of structures (which may affect how repairs can be carried out), the need to keep operating until repairs made out, limited time available to complete repairs or the particular requirements of given operating environments are all often better understood by an engineer specialised in structural integrity.</li>
<li>Development of inspection regimes requires experience of carrying out detailed inspections to know where to look and how to look.</li>
</ul>
<p>For these reasons ensuring structural integrity of structures requires structural engineers who specialise in structural integrity.</p>
<p>This should not be meant to imply that a specialist in structural integrity is all that is required to maintain structural integrity. For example, when assessing whether a structure has appropriate integrity, it may be necessary to engage a whole range of additional expertise. The original designer knows what assumptions were made in the design of the structure. Experts in advanced structural analysis, fatigue behaviour, corrosion chemistry or geotechnical behaviour may all be required. Experts in construction may be required to assist in repairs, and if the structure no longer has enough integrity remaining experts in demolition may be required.</p>
<p>However engineers who specialise in structural integrity bring experience of inspection, knowledge of deterioration and understandng of how structures are used in the real world that is essential for maintaining structural integrity. This is particularly the case where unusual structures, high loads, harsh environments and fast inspection / repair turn-around requirements are present.</p>
<h2 id="conclusion">Conclusion</h2>
<p>Structural integrity is a fundamental part of ensuring that structures do what they are designed for, and are safe when conditions exceed the original design intent. However, maintaining structural integrity is not simply a matter of good design. Structures require inspection and maintenance throughout their life to ensure they continue to have appropriate levels of structural integrity. Engineers who specialise in structural integrity play a vital role in ensuring that structures are safe and reliable throughout their life.</p>Sean KaneThe company I work for specialises in Structural Integrity, which is a niche area of Structural Engineering. At university I never even heard the phrase “structural integrity” let alone learnt anything about it. A google search shows a few references to structural integrity in a mechanical or aeronautical context, a lot of engineering companies claiming to provide structural integrity services and not much else. A search on Amazon shows that there isn’t even an introductory textbook.Hello World2018-09-28T00:00:00+00:002018-09-28T00:00:00+00:00http://ifitsmoving.com/2018/09/28/hello-world<p>Welcome to <em>If it’s Moving, it’s Broken</em>, a structural engineer’s take on the world of of Structural Engineering, Structural Integrity, Computer Programming and anything else that interests me. I am hoping that by writing this blog I can share a little of what I do, and at the same time improve my writing ability.</p>
<p>The title: <em>If it’s moving, it’s broken</em> comes from a joke describing the difference between structural and mechanical engineers. For structural engineers, if something moves it usually means that large amounts of steel and concrete will soon be coming towards you at high speed. The mechanical engineers I work with though usually want things to move - if they don’t it probably means you’ve had a bearing fail.</p>
<p>However, in the real world it’s never that simple. Bulk materials handling machines that load mineral products (such as conveyors, shiploaders, stacker-reclaimers etc.) are very large structures that move. They don’t fit neatly into either structural OR mechanical engineering. Understanding this intersection is an interesting challenge that this blog may explore.</p>
<p>In addition to structural engineering, I am interested in programming and how it can be relevant to the engineering world. Currently the area of structural engineering is both heavily innovating with computer analysis tools (i.e. Rhino + Grasshopper for parametric CAD modelling that is then often placed straight into analysis packages with their APIs) but also lagging behind in many ways. For example, once design of members is done, design of connections and local details is often still a very manual process, and spreadsheets abound (many of poor quality).</p>
<p>This is an area that I am exploring with my own dabbling in programming. Given I work full time as an engineer, I don’t have time to waste on curly brackets, and therefore <code class="language-plaintext highlighter-rouge">Python</code> and (if I must) <code class="language-plaintext highlighter-rouge">VBA</code> are my languages of choice.</p>
<p>Hopefully this post gives you an idea of what my goals are in writing this blog, and an idea of whether or not you’ll be interested. I hope to see you around.</p>Sean KaneWelcome to If it’s Moving, it’s Broken, a structural engineer’s take on the world of of Structural Engineering, Structural Integrity, Computer Programming and anything else that interests me. I am hoping that by writing this blog I can share a little of what I do, and at the same time improve my writing ability.