MANUFACTURING HEAT TREAT

Welcome to another episode of Heat Treat Radio, a periodic podcast where Heat Treat Radio host, Doug Glenn, discusses cutting-edge topics with industry-leading personalities. Below, you can either listen to the podcast by clicking on the audio play button, or you can read an edited version of the transcript. To see a complete list of other Heat Treat Radio episodes, click here.

Audio: Heat Treat Modeling With Justin Sims

In this conversation, Heat Treat Radio host, Doug Glenn, interviews Justin Sims of DANTE Solutions about heat treat modeling. As the heat treat world moves farther way from mysterious black box processes, find out how the latest advances in heat treat simulation software can help your company model specific processes and materials in advance, leading to less guesswork and more profit.

Click the play button below to listen.

Transcript: Heat Treat Modeling With Justin Sims

The following transcript has been edited for your reading enjoyment.

We're going to talk to Justin Sims, lead engineer at DANTE Solutions, Inc., about heat treat modeling.  It's a pretty interesting topic.  With all the advances and sensors and computing power, the heat treat world is moving further and further away from the mysterious black box processes of yesteryear and is allowing companies to model specific processes and specific materials in advance so that there is less guesswork and more profit.  DANTE provides the means by which companies can accurately predict what is going to happen to their part during the heat treat process.

DG: Justin is not only the lead engineer at DANTE Solutions, he is also the author of an article that just appeared in the March 2020 issue of Heat Treat Today and the title of the article was Process Innovation To Reduce Distortion During Gas Quenching. It was a pretty interesting article, something worth reading if you haven't already.  It has to do with DANTE controlled gas quench.

JS:  I got my bachelors in mechanical engineering degree from Cleveland State.  I graduated back in 2015.  I actually started interning at DANTE in 2014 and went full-time in 2016.  I've been the lead/principal engineer at DANTE with mainly responsibilities of managing projects, training our DANTE users, and offering support to our DANTE users.  I helped develop our patent-pending DANTE controlled gas quenching process, which you had just mentioned, and then also a little bit of IT, marketing, sales, and shipping. Being a smaller company, we can kind of do it all.

DG: Tell us briefly about DANTE.

JS: DANTE Solutions is an engineering consulting and software company.  We offer consulting services as well as licensing our software.  We mainly focus on the aerospace industry, the auto industry quite a bit as well, and we've been starting to get into the mining and energy sectors also.  As I said, we are a smaller company.  There are six of us right now. Two to three guys mainly focus on the software side, and the rest of us focus on more of the training, the support, and the consulting side of the business.

DG: DANTE is located near Cleveland, OH, and Lynn Ferguson, who has been in the heat treat industry for many, many years, was one of the founders. Let's talk about the genesis of the software. Would you say the software is the core product that DANTE Solutions offers?

JS: Yes, it is. We mainly stay in consulting to stay current and to give those users who don't have the capability to run our software (either they don't have the hardware or they don't have the analysts to be able to do such a thing), so we still offer our consulting services for them. But mainly, software is our main line of business.  DANTE was actually formed back in 1982 as Deformation Control Technology, Inc., and we changed our name in 2014 to actually reflect more of the software side, so that's when we changed to DANTE Solutions, Inc.

The project itself that DANTE came out of actually started in 1994 and 1995.  It was a collaboration between Ford, GM, Eaton Corp. and then four national labs--I believe they were Los Alamos, Sandia, Oak Ridge, and Lawrence Livermore--and then us as Deformation Control Technology.  The whole project came out because those large automakers were claiming millions of dollars of lost scrap from distortion.  It was starting to become a major issue and they wanted a way to be able to model the process and be able to optimize the process a little bit better.  After that project ended, DANTE somehow ended up with the software, which has worked out well, as we've been able to commercialize it and we've been updating all the material models and the material database for the last 20 years.  It's actually come quite a long way.

DG:  How did you segue over from auto industry into aerospace?

JS:  It just happens that the aerospace components cost a whole lot more than the auto industry components.  It was a natural fit once they realized that this software was viable and could do what they needed it to do.  And aerospace seems to be more receptive to modeling because their parts are so expensive.

DG:  Let's try to put a little flesh on the bones here. For a manufacturer who has their own in-house heat treat for aerospace, automotive, energy or whatever, what makes this software attractive?  What makes it viable?  Why would someone want it, and why and how do they use it?

JS:  Let's start with viability.  The first thing is that it is easy to use.  DANTE is a set of material routines that link with Abaqus or Ansys finite element solvers.  These are solvers that engineers and analysts in the industry already know pretty well, so there is not a lot of learning of new software.  DANTE is just a material model, so all you're really responsible for is the material name and what microstructural phases you're starting with.  Then we have the ability to modify a few of our control parameters, activating different models; we've introduced stress relaxation, carbon separation, carbide dissolution, and all these different models that you can activate.  But the biggest thing that trips people up . . . [is] understanding your process.  We like to work with people a lot on trying to help them understand what type of thermal behavior their processes are actually imparting on components.  We've done a lot of work with setting up their essentially quench probes and be able to turn around and be able to take that back to heat transfer coefficients that get put into the model. As far as DANTE is concerned, it is fairly easy to use.

We've also developed what they're calling ACT (Ansys Customization Toolkit).  It is essentially a series of buttons where you would click on these buttons, fill out information, and then essentially run your models. Abaqus, for the new version of DANTE, we've also developed a plug-in that essentially does the same thing.  So DANTE has become very point-and-click.  In this world, I think people like that simplicity.

The next big one would be the accuracy that everybody is concerned about.  Our accuracy is due to the models that we use and the algorithms that we employ.  There are two types of accuracy.  I've touched on the boundary condition accuracy, and that is how your process behaves thermally.  That accuracy can be tough to get.  It's very doable and we've helped people achieve some really amazing accuracy.  The relationship I like to use here is people know static loading models and a lot of engineers have run static loading models.  The loads that you put on these static models are going to determine what deflections you get.  If your load is not correct, then your deflection will not be correct.  In heat treat modeling, the thermal boundary condition is your load.  The more accurate your heat transfer coefficient can be, the more accurate your results are.  But, with that being said, you can still gain a lot of valuable information from being close enough.  We'll talk a little bit about that with the uses and whatnot.

The first important model type that we use is the mechanical model.  We use a multiphase internal state variable model.  A conventional plasticity model considers stress as a function of strain only, where the internal state variable model actually accounts for the history of deformation by relating the stress to dislocation density.  It actually accounts for the history of deformation, which is very important as the steel goes through all the stress reversals that it does going through the process.  Our mechanical model defines each phase, so austinite, pearlite, ferrite, bainites martensite, tempered martensite, all of them, as a function of carbon, temperature, strain and strain rate.  It also accounts for the trip phenomenon.

For our phase transformation model, we like to use analytical models instead of TTT CCT diagrams, and we do this because you don't get any transformation strain information out of the diagram.  So you have no idea how much it is deforming.  In order to figure that out, we like to use dilatometry tests to fit to our analytical models.  We also account for carbide growth and dissolution during carburizing, which is becoming a major point of interest due to the high alloy content of some of these steels that they're now trying to carburize.

DG:  Let's talk a bit more about where manufacturers, who have their own in-house heat treat, might use DANTE's software tool.

JS:  One of the big things we like to use it for is what we call sensitivity analysis.  This would be, "what happens if my normal process has a little bit of variation?"  Or, "what happens if my process parameters change a little bit?" We've also worked into the model now normal material variation.  So if your alloy content is a little on the high side, how would the material behave?  If it's a little on the low side, how will it behave? [This] is a big deal.  One example would be, "I just designed a new part and I want to make sure that it behaves given the range that I know my process can vary."  All processes will vary.  This is no way to make the process exactly the same every time.  Also, in the sensitivity, you can ask the question, “What process variable is a distortion or stress most sensitive to?"  By finding out what process variables cause the most sensitivity, then those are the process variables you really need to pay attention to during processing, then the other ones you can just make sure they're in range and leave them alone.

Development and design are two of the big ones that we're trying to get out there that this software can be used for.  Everybody knows that it can be used for troubleshooting.  Once something goes wrong, yes, sure the software is great and we help figure out a problem; but why not find the problem before it ever even happens?  We've been trying to get people to use it for development of new carburizing and nitriding schedules as well as new recipe and design, and even novel processes.  You had mentioned our DANTE controlled gas quench.  That actually was conceived through all the modeling that we do and watching the response of the material and saying, “Wait a second.  If we can control the martensite transformation rate, we can really control the distortion, so let's see if we can do this.”  Things like that can come out of the software.  Design as well, of optimizing shapes for quench.  You can even do quench to fit, which is, "I know my part distorts this much, so let me machine it distorted and then it will fall into shape."  Optimizing processes.  All of that can be done through design development, and you can find these problems before they ever happen.

Another really big one that I like, and Lynn, our owner, is really keen on this one, is the understanding of your process.  When you start to set up these models, you have to ask a lot of questions about your process.  What is the HTC of my process, which relates back to agitation in the tanks, part racking, flow directions?  You really need to know times and temperatures of every step in your process.  So not just the heat to quench, but what about all those transfers in between?  All of that needs to be done.  So you end up asking a lot of questions like that.

The other one that I always like to say is that the heat treat software removes the black box.  In the past, you know what goes in and you know what comes out, but what happens in the middle is kind of a mystery.  The software helps you figure out what exactly goes on during your process.  It can be very eye-opening.

DG:  I've talked with James Jan and Andrew Martin over at AVL, and we talked about a variety of ways they use some of their software, and they mentioned that they work with you guys as well, and they were talking about not even just like a quench agitation, flow direction, and things of that sort, but part orientation as it goes into a quench.  I assume that would be something also that you guys would be able to help analyze, right?  Which way to even put the part into the quench?

JS:  Sure, sure.  And we've done that.  The one that comes to mind is a long landing gear.  This landing gear was about 3 meters in length, and we looked at even slight angles going into the quench tank can have serious consequences on the distortion.  That is definitely something that we've looked at in the past.

DG:  Just that orientation would help, but maybe eliminate vapor stage, or whatever, I assume?  Or pockets?

JS:  Right. And even beyond that, it sets up thermal gradients in different locations of the part.  So now instead of cooling one section faster, you're cooling it a little slower and that kind of thing.  That also relates back to actual vapor stages and how bubbles get trapped.  But that goes back to defining boundary conditions, which is where software like AVL's FIRE can really be helpful in understanding flow patterns.  There is a beneficial relationship there.

DG:  There are a host of different materials that people are using.  How broad is the database, as far as the different types of materials, that you can analyze and model?

JS:  That is a good question.  We have a lot of low alloy, medium alloy, and carburizing grades of steel, the 1000 series, the 8600 series, 9300 series, those types of materials.  We've also worked with some of the high alloy aerospace grades like C64 and the Pyrowear 53 and that sort of thing.  But right now, it's all steel.  There is a lot of talk about being able to do aluminum.  We get that question a lot.

DG:  I was wondering about that specifically- aluminum and/or of course, when we talk aerospace, we're talking titanium.  So titanium is not on the table at the moment?

JS:  It is, but it isn't.  The interesting thing is that there is a phenomena precipitation hardening that goes on in aluminum and titanium.  But it also goes on in these high alloy steels.  It is a secondary hardening mechanism.  We've been working on that and we feel that once we can handle secondary hardening in steel, then the jump to aluminum and titanium should be pretty straightforward.

DG:  So to recap, for those of us who are not as well-versed in the product as you are, basically you've got a simulation software that takes into account the material that is being used, also the thermal process (the recipe), which would include both a controlled heat up and potentially a controlled quench.  Is that a reasonable way to describe it in a very broad way?

JS:  Yes. And also, even the steps before that, like carburizing.  If the part is carburized, you would carburize it first.  Or nitriding; we've just introduced those models.  You can literally do the entire process.  And it's not just quenching either.  We've done martempering, austempering, normalizing, all of these things.  Most all normal thermal processing, DANTE can handle.

DG:  The last question I want to ask is, Who is the ideal person/company that would really find the product/service that you're providing useful? I know you mentioned aerospace and automotive, but can we be more specific than that? Where are you finding the most success?

JS:  That's a tough question. Generally, everybody that has used our software has found real benefit in it.  We've tried to get testimonials from a lot of folks, but this can be difficult because of their companies.  But from Cummins, we've gotten good responses and also from GM we've gotten good responses.  One of them has used it to actually introduce new material and replace legacy material that is now saving them quite a bit of money.  GM has used it to look at process design and optimization.  But I would say mainly the people that are going to benefit the most are the folks that have an analyst to be able to do the simulation almost on a daily basis.  It's one of those things where the more you do, the more you see and the more you understand what is happening.  But really anybody that does heat treatment can benefit from understanding what's going on in their process.

DG:  You mentioned Cummins, and I'm looking at your website, and I just want to read a paragraph:

DANTE heat treat simulation software has been a great boon to Cummins.  Since we've started using their software, we have gone through several projects that have increased our understanding of heat treatment and some of which have saved us production costs.  One example was enabling us to gain the leverage needed to make a material and process change on a legacy product that is now saving us at least 25% on material costs.  The team at DANTE Solutions has always been very accommodating and is very quick to give assistance and feedback whenever troubles arise, even when the troubles are caused by other parts of the simulation and not DANTE itself.  I look forward to working with DANTE team in the coming years as we expand our list of engineers who use this software. -- Brian W. at Cummins

So that leads me to one other question.  When a person interacts with you, are they buying software as a service?  Is it cloud-based or is it something that they purchase a license for one computer, one user?  How does it work?

JS:  There are a couple of different ways.  They can lease it annually or they can essentially buy the software and lease a license annually.  The software can go either on their computer or it can go on a server at their company.  We also have options for corporations where you can essentially get software at different locations.  We have a lot of options and we can work with customers if they [have] unique needs.  That's one of the benefits of being a smaller company, we're pretty flexible like that.

DG:  DANTE's mission statement from their website has a nice ring to it: “DANTE Solutions is determined to promote the use of simulation in the heat treat industry.  From design to troubleshooting, DANTE Solutions believes everyone can benefit from a little simulation in their life.”

If you'd like to get in touch with Justin Sims at DANTE, please email me, Doug Glenn, directly at doug@heattreattoday.com and I'll put you in touch with Justin.

To find other Heat Treat Radio episodes, go to www.heattreattoday.com/radio and look in the list of Heat Treat Radio episodes listed.