EXPLORE THE POSSIBILITIES
Where’s your analysis?
Adding corrosion mitigation to your digital thread
I must confess that corrosion isn’t something I know much about. Nor is it something I’ve given much thought to, besides booking an appointment for my car’s annual rust-protection treatment. However, since we’ve started making things with metal, corrosion has been there to foil our plans. Sure, there are tried and tested ways to control the spread of it. However, it didn’t strike me as a pressing engineering problem seeking innovation, just a necessary step for durability and safety. It was something I didn’t associate with simulation, let alone digital transformation or shifting left.
That was all before I met Dr. Alan Rose, CEO of Corrdesa, LLC. Before we get into who he is and what his company does, he shared an anecdote that reinforced my misguided perception:
Corrosion control has traditionally been kind of slapped on or bolted on at the end. I had a friend of mine who worked at a helicopter company, so there are not many of them around. And he'd tell me a story how all the engineers would meet together, and you'd have your thermal engineer, discussing some change and the thermal engineer goes, oh no, you can't do that. It'll get too hot. The aerodynamicist would say, no, you can't make that change it'll be too heavy. The structural engineer would say, oh, you can't do that change. The structure wouldn't take it.
And they'd all have their FEA analyses and the color maps and so on. And then they'd come to my friend who was a materials guy and he’d say, no, you can't do that, it will corrode, or the materials are incompatible. And they'd all look at him and go, where's your analysis?
So, what exactly is galvanic corrosion? According to everyone’s favorite AI chatbot, it is a form of corrosion that occurs when two dissimilar metals come into electrical contact with each other in the presence of an electrolyte. One metal, called the anode, corrodes preferentially and loses electrons to the other metal, called the cathode, which corrodes less rapidly. This process can accelerate the corrosion of the anodic metal, and can also protect the cathodic metal from corroding.
Corrosion analysis looks at the impact on materials and equipment. This can include the examination of chemical properties and environmental factors. The goal is to identify the root cause and develop strategies to prevent or mitigate it. Sounds easy enough.
However, corrosion is a slow process, and physical testing can take months to years. Quite the obstacle to overcome in an age where the opposite is expected – shorten design and test cycles. Ideally, by shifting left and getting ahead of issues earlier on in the process, we could achieve both objectives. Is that possible?
Alan and I discussed this in detail. It’s hard to say if he is more knowledgeable or passionate about the topic. Likely both in equal parts. He illuminated to me how, with simulation, the archnemesis of metal can be put into check.
So why is corrosion such a big issue?
Why is corrosion a big issue? Well, we have cost, so it's about two and a half trillion dollars of cost globally. So that's like what, about 3% of global domestic product? There's a cost. There's obviously safety, and it takes a long time to test materials and deploy your new coatings and new designs. So, you need to be able to shorten that.
On a cost perspective, we do a lot of work for the Department of Defense, corrosion there costs like $22 billion a year, and about half of that is for aerospace.
So no matter which way you cut it, there's a lot of money to be saved in quickly assessing and modeling corrosion as part of your design process. Just like thermal analysis, stress analysis, I think now computational corrosion analysis is kind of the new discipline.
If it is such a big issue, then why haven't we seen this discipline popping up in more places? What are the challenges when it comes to implementing corrosion simulation?
Compared to thermal analysis and stress analysis, the biggest difference is the necessary data has been already collected and is available. When you use Simcenter STAR-CCM+, you can very easily find in a database, thermal conductivities, Young's modulus, and so on. But for corrosion analysis, the data just simply isn't there.
Corrdesa has designed a protocol, in our partnership with the United States Navy, on how to acquire that data, and now that's part of a standard. So, people know how to acquire this data and how to actually insert it into Simcenter STAR-CCM+ to do the 3D corrosion calculations.
Now we have not only the tools but also the validated and robust data that enables this technology to move forward.
Physically testing for corrosion can take weeks, months, or years. The elapsed time required to validate results means It's usually deemed too big of a risk to implement new coatings or materials. Tried and tested materials win out because they’re a known quantity.
Alan doesn’t disagree that traditional methods are risky. But he argues that simulation can help companies explore possibilities without jeopardizing their project timeline.
Implementing new materials is definitely a challenge, certainly in aerospace. That's my background, in aerospace. Some argue the best test for materials is actually in the field, but with the pressure to shorten design cycles, can't afford that time. By that I mean putting your sample on the side of a plane or a ship or a car and drag it around Michigan. That can take months or years.
So, then they have beach tests where you put material samples in a corrosive environment on the beach, say down at Kennedy Space Center. That can still take six months. Our accelerated corrosion chamber tests take a month.
if you've got a large project and you're thinking I really need the best and latest in materials, you've got to consider that as a risk. It might turn out that the materials aren't appropriate. So, it's back to the drawing board.
But we've developed techniques that can within say about 24 hours you can characterize the materials in a laboratory, laboratory scale. You can then put that material information in your computer model, and you can actually model the corrosion within minutes to a few hours, depending on the complexity of the geometry. There’s a lot less risk now. You can look at many, many other material options rather than just try and figure out what you can do in the time given to you.
Lessening the risk in deploying new materials
If there was less risk in deploying new materials, there is a greater chance they would be used in large projects, leading to more innovation. He elaborated with an industry example:
“An example is automotive. They must go from a blank to market within, say, two years. It takes about 12 months to design the car. That only gives you 11 months to get your tooling done and all the tests. You can only do one full-scale corrosion test, perhaps two. If you can do all this computationally, then that will reduce the risk significantly.”
Alan goes on to elaborate that changes in regulations, which requires finding alternatives to toxic materials that have been mainstays, add even more pressure.
“Traditionally we've been able to use some materials such as cadmium and chromates that are very effective in mitigating corrosion. we find ourselves having to swap out those materials and drop in new materials that are a lot less toxic to our operators and to the environment. Now we've had four or five decades to understand some of these older materials, but we are forced to implement these new materials in shorter time scales, in about five years or so because of legislation like REACH and OSHA and so on.”
But it’s not just automotive. 80% of aircraft structural failures are from corrosion pits. Can simulation predict and prevent that?
Yeah, so in teardowns of aircraft, there's been reports have shown that about 80% of structural failures can be sourced back at corrosion pits. The choice of materials and repair methods and sustainment is very important.
Implementing computational corrosion techniques can very, very quickly help you do that by basically getting your result within a few hours.
Who exactly is using all these corrosion prediction techniques?
There are kind of three parts to that. We had to consider what the user personas would be. Now computerized engineering traditionally would require somebody with expertise that knows how to create the meshes for analysis, like in computational fluid dynamics because that's what we're using.
Do you train your materials people to be CAE experts, or do you train your engineering team to be materials experts? You can't do that. We took these personas, the CAD people, computer-aided engineering people, and materials process people, and developed a toolset for their respective needs.
You can do what we call our corrosion Djinn analysis -- that just looks at materials. Have I got a problem here? Is there going to be some issues or incompatibility? And that can highlight on your system where you may have some issues.
Then you might want to drill down and get more details on the geometry. If you have somebody that's familiar with CAD, that person could just take that whole system and then model and predict the corrosion on that 3D assembly.
But if you have a materials process person (MPP), they don't have that expertise. So, we've created a GUI within Teamcenter, where under the hood it runs Simcenter STAR-CCM+ automatically using common fundamental building blocks. What do I mean by that? If you think about it, an airplane is made of lots of nuts, bolts, lap joints, butt joints, etc. We've captured those fundamental geometries.
Using that, the MMP can create the mesh and set up the analysis without being a typical CAE user. It then reports back to the MMP guy, the corrosion rates, and the map of the corrosion rate around the whole assembly or sub-assembly. So
Can you build corrosion control from the outset?
That's a good question because that's certainly what the Department of Defense would like. There are standards and guidelines, and requirements, it's called corrosion prevention control planning. And that takes a lot of information. You need to know the materials, the data, and the geometry, and that can all be kind of accounted for and controlled in an environment like Siemens Xcelerator.
Corrosion modeling in Siemens Xcelerator makes it possible to build in corrosion control and compliance from the outset. When you're looking at requirements through to surrogate designs and testing and managing all the data all the way through and predicting the corrosion and durability of that product.
What’s next for Corrdesa?
Our USAF project for creating the tool, that is ‘The Corrosion Modeling Toolset’, finished its first phase in late summer 2022. It has been demonstrated to several maintenance groups responsible for several USAF platforms. This has created a lot of traction and there was considerable interest at the 2022 Aircraft Structural Integrity Program (ASIP) Conference.
We are now in a second phase where we are connecting actual field results from aircraft with simulations from the toolset. Furthermore, we will be implementing a module that will account for changes in the environment and aircraft mission by considering weather data.
In response to traction from the automotive community, Corrdesa is presently working on a project to gather materials and coatings data relevant to automobiles, this will result in about 40 further materials added to the Corrosion Djinn database, thereby extending this capability to Automotive designers and materials engineers.
Corrdesa develops and uses computer-aided engineering (CAE) simulation tools to identify and mitigate corrosion risk by optimizing and implementing corrosion-resistant coatings, processes, and equipment ensuring compliance with regulatory and environmental requirements.