FreeCAD News

Peo­ple have been using FreeCAD in acad­e­mia and research for many years, and the range of these research projects is sur­pris­ing­ly broad.

Today, we are talk­ing to Matt Thomas aka LT72884, who has a great YouTube chan­nel where he explains how to use free and open-source engi­neer­ing soft­ware for rock­et design, CfD sim­u­la­tion, FEM, and data analysis.

Hi, could you please briefly intro­duce yourself?

Yeah, sure. My name is Matt Thomas, and I am very pas­sion­ate about sci­ence and engi­neer­ing. I grew up with a moth­er who was a head nurse and team lead for the first heart trans­plant, as well as a team mem­ber of the Human Genome Project, so sci­ence and med­i­cine were a very com­mon din­ner table con­ver­sa­tion for me grow­ing up.

My dad, on the oth­er hand, was a telecom­mu­ni­ca­tions spe­cial­ist for the mil­i­tary, and he would bring me along for some of his jobs, such as installing very large phone sys­tems for the base and the research facil­i­ty at the local uni­ver­si­ty.

When I was in high school, I learned about com­put­er net­work­ing and web design because I was hor­ri­ble at math and sci­ence. I fell in love with tech­nol­o­gy, and after grad­u­at­ing in 2003, the local uni­ver­si­ty had a sis­ter pro­gram with Carnegie in Net­work Engi­neer­ing and cyber­se­cu­ri­ty. I received my CCNA, CCNP, RHCT, Eth­i­cal Hack­ing, and a few oth­er well-known cer­ti­fi­ca­tions, as well as a Bach­e­lor’s degree.

How­ev­er, when I was claim­ing grad­u­a­tion, some­thing just didn’t feel cor­rect. So, I had a long and excel­lent con­ver­sa­tion with my moth­er, who has nev­er led me down the wrong path. She sug­gest­ed that I have always liked sci­ence and tech­nol­o­gy, so maybe there is a way to com­bine the two. So I did just that. 

I was 26 years old when I decid­ed to go back to school for med­i­cine and engi­neer­ing, the only prob­lem was… I had 3 years of basic math to make up before I could even apply for the asso­ciate of pre-engi­neer­ing and my micro­bi­ol­o­gy coursework. 

After 3 years, I final­ly made it into the pro­gram and con­tin­ued to push through 6 more semes­ters of math, 3 semes­ters of chem­istry, biol­o­gy, patho­phys­i­ol­o­gy, and micro­bi­ol­o­gy. Then I trans­ferred to anoth­er uni­ver­si­ty where I fin­ished my degree in Aero­space engi­neer­ing with a mas­ters course­work in Com­pu­ta­tion­al Flu­id Dynam­ics (CFD), numer­i­cal and data analy­sis, as well as advanced mate­r­i­al sci­ences. 

This is your sec­ond high­er edu­ca­tion, right? Why did you choose engineering?

To be hon­est, I chose engi­neer­ing to hon­or my fam­i­ly and mom. I grew up in a Chris­t­ian home, so I was always curi­ous how the big guy in the sky orga­nized things. I con­sid­er him to be the ulti­mate engi­neer and sci­en­tist, as did my moth­er.

Engi­neer­ing and sci­ence help me under­stand how things in this world and out of this world work. I love to just know how things oper­ate, even if I nev­er use that device or prod­uct. Oh, and the world is not flat 🙂

How many papers have you already pub­lished where free and open-source soft­ware is used prominently?

I have about 8–10 papers using open source soft­ware as the pre­dom­i­nant appli­ca­tions, such as QBlade, Open­FAST, MLife, Octave, FreeCAD, Ond­sel Engi­neer­ing Suite, Open­FOAM, Par­aView, Pre­Po­Max, and Fat­lab (an open-source fatigue analy­sis tool), with ver­i­fi­ca­tions done with ANSYS.

Many of the papers deal with wind tur­bine dura­bil­i­ty and fatigue analy­sis (“Esti­mat­ing Wind Tur­bine Blade Dura­bil­i­ty Uti­liz­ing Nation­al Renew­able Ener­gy Lab­o­ra­to­ry Tools” is the lat­est) and can be found on Research­Gate. Some of them may have sim­i­lar titles and infor­ma­tion, but they are exten­sions of our orig­i­nal work, mean­ing, we pub­lished a paper, then com­plet­ed even more work, and added to it for a sec­ond pub­li­ca­tion and peer review. I even wrote a basic 75-page user man­u­al for Open­FAST so that any­one want­i­ng to learn how to use it can. 

Why is it impor­tant for you to rely on free/open-source tools?

First, they are avail­able to research stu­dents and interns, or any­one, for that mat­ter, for a real­ly good price. Sec­ond, open-source engi­neer­ing tools are very impor­tant, espe­cial­ly for devel­op­ing nations that can not afford the dras­tic price tag of some pack­ages. How­ev­er, peo­ple are skep­ti­cal of using open-source soft­ware because they are afraid it’s “unsafe” or will pro­duce inac­cu­rate results. This is why hav­ing NASA per­form val­i­da­tion tests with Open­FOAM was such a huge deal for the open-source community.

How did you get involved with rockets?

It was an intern­ship and practicum at school. We were tasked with cre­at­ing an active drag sys­tem for a rock­et to achieve a spe­cif­ic alti­tude. They are like air brakes in a sense, and from there, I just kept learn­ing more and more about rock­ets, which led to for­mu­lat­ing and char­ac­ter­iz­ing my own sol­id rock­et fuel using ammo­ni­um per­chlo­rate, met­als, and epox­ies, the same ingre­di­ents they use for much larg­er rockets. 

And then you got hooked?

I have always loved planes, and I often go to the aero­space muse­um that is close to my home.    When the school offered me and a few oth­ers this chance, I took it and have loved it ever since. Mak­ing my own rock­ets, fuel, and see­ing them go from dig­i­tal con­cept to real­i­ty is very rewarding.

I’ve also noticed that wind tur­bines and their dura­bil­i­ty are a big deal for you. How did the inter­est originate?

Some­one from the state research depart­ment on wind ener­gy read my design the­sis on high-pow­ered rock­ets, which includ­ed a lot of test­ing and analy­sis. So they asked me if I want­ed to join the team for a few months as a full-time job doing fatigue analy­sis and design for wind tur­bines. I said, “Sure, why not?”. The job last­ed 2 years because they real­ly liked my work and how I han­dled things.

You use a whole vari­ety of tools: Open­Rock­et, FreeCAD, the Cfd­Of work­bench and Open­FOAM, Par­aview, Open­FAST, MLife, and even Excel. That’s a lot of bridges to build between appli­ca­tions. How well does it work for you, and where do you think are the worst inte­gra­tion bot­tle­necks today?

So far, I have not had any issues exporting/importing data from one appli­ca­tion to the oth­er. Par­aview plays very nice­ly with Open­FOAM data, and the inte­gra­tion of Par­aView into the CfD­of work­bench for FreeCAD has been nice, though I pre­fer to close FreeCAD after every CFD sim­u­la­tion and open the data into a fresh start of Par­aview.

I actu­al­ly use pyDatView from GitHub as a way to view some of the Open­FOAM data files. It’s an old tool that I just discovered.

Open­FAST and MLife both export the data as Excel files, so that makes it very nice. I do not use Libre­Of­fice Calc as it is miss­ing a few major func­tions need­ed for most engi­neer­ing work­flows. This is why I use Excel. Their solver and numer­i­cal meth­ods are excellent

Open­Rock­et is one of my favorites, but it is a stand­alone soft­ware. It will cal­cu­late the forces and drag coef­fi­cients for you based on the Bar­row­man Equa­tions devel­oped by James Bar­row­man at NASA. Using just geom­e­try, the Reynolds num­ber, and skin fric­tion val­ues. Open­Rock­et will allow you to graph many val­ues against each oth­er, mak­ing it a very valu­able tool if CFD is not an option. 

Now, speak­ing of CFD, I have set up many sim­u­la­tions (again, thanks to the forums for help­ing me get the pre­lim­i­nar­ies set­up) of high-pow­ered rock­ets to test against Open­Rock­et, and sure enough, I get pret­ty dang close to the same values. 

Most peo­ple, when they do CFD for the first few times, get results that are way off, and then they get dis­cour­aged. 95% of the time it’s because they have too coarse of a mesh and not enough refine­ment areas for the mesh to accu­rate­ly pro­duce good results. This is why I made the YouTube series on CFD with Ondsel/FreeCAD. 

Well, my pri­ma­ry rea­son is so I don’t for­get how to do it myself. I record a lot of things I do so I can ref­er­ence them lat­er, which comes in handy if a super­vi­sor asks about a process.

As for bot­tle­necks, Open­Rock­et, an advanced sim­u­la­tion tool, is much hard­er to use and inte­grate with oth­er soft­ware. Pre­Po­Max actu­al­ly works bet­ter as a stand­alone prod­uct, i.e., import the STP file, mesh it, then set up the sce­nario, all in Pre­Po­Max. That is one bot­tle­neck. The oth­er is that FreeCAD needs to improve the mesh­ing defaults in the FEM work­bench. Pre­Po­Max has killer defaults and mesh­es extreme­ly well, we just need that in FreeCAD now.

When we first dis­cussed your aca­d­e­m­ic work, you said that the paper that you pre­sent­ed at the IEEE con­fer­ence in May 2024 was well-received, espe­cial­ly from oth­er coun­tries and devel­op­ing nations, and that was your focus. Could you please elab­o­rate on the focus part?

For sure : ) My main focus at that pre­sen­ta­tion was to spread the word on how reli­able FreeCAD, Open­FOAM, Open­FAST, and MLife are for major engi­neer­ing projects. I mean, I had been doing stress-test­ing and design­ing 65-meter to 100-meter wind tur­bine blades in FreeCAD with ease(thanks to the advice from the forums when i got stuck), and then per­form­ing all CfD, FEM, and stress analy­sis with FreeCAD and OpenFAST/MLife. 

Being a devel­op­ing coun­try and feel­ing lost in how to begin projects due to high-cost soft­ware is very stress­ful. I have want­ed to share this knowl­edge, and why not share it at a very large IEEE event?

As soon as I fin­ished my pre­sen­ta­tion, I had 3 or 4 atten­dees come over and ask for copies of all my papers to take back to their coun­tries to share with offi­cials they knew. 

Have you been in touch with them ever since?

I have tried a cou­ple of times over the last year, and have not heard any­thing back. Could be due to inter­net laws in their coun­try, which could pre­vent communication.

When you talk to your mates at uni­ver­si­ty and to your teach­ers, do you see much antag­o­nism towards free and open-source soft­ware? If you do, where do you think it is com­ing from?

Yes and no, but most­ly no! My pro­fes­sors encour­age the use of open-source soft­ware to teach us how to val­i­date our results. We then share our own per­son­al results with about 30 oth­er stu­dents and 4 oth­er pro­fes­sors for review. It’s an excel­lent way to teach new stu­dents the impor­tance of peer review and safety. 

I think the biggest frus­tra­tion with FreeCAD that users and uni­ver­si­ty stu­dents have is the workarounds that we have to do often to get 3 com­mon fea­tures to work, such as thick­ness, fil­let, and cham­fer, but that is an issue to do with the ker­nel, which FreeCAD does not man­age unfor­tu­nate­ly.

My flu­ids pro­fes­sor uses Open­FOAM and prefers it over oth­ers because it requires stu­dents to under­stand the the­o­ry of flu­id dynam­ics rather than just push­ing but­tons. Any­one can push but­tons and get data back, but, like he says, “Garbage in is garbage out”, or GIGO for short. I still pre­fer to push but­tons, I know the the­o­ry, and now I like to just push a but­ton and let the PC do all the work.

Where the antag­o­nism comes from is the data sci­ence group, and I don’t blame them. As I pre­vi­ous­ly men­tioned, Libre­Of­fice Calc and oth­er open-source spread­sheet pro­grams lack some major func­tions, such as a non-lin­ear solver that actu­al­ly works with­out wast­ing time with some sort of workarounds. 

What’s the next big chap­ter for you research-wise?

In the last 4 years, I pub­lished a few papers on food-safe 3D print­ing. I’m cur­rent­ly doing the last exten­sion of the orig­i­nal paper, test­ing fil­a­ments for tox­ins and heavy met­als via leach, migra­tion, and mass spec­trom­e­try testing. 

My first pub­li­ca­tion on this mat­ter illus­trat­ed that lay­er lines do not pose a risk and can be cleaned and san­i­tized to safe lev­els, and that noz­zles do not leave any lead either.