S2E55 Greg Trees | FEA, EHD, and Other Useful Engineering Acronyms

Greg Trees has a degree in mechanical engineering, and has spent nearly 30 years as a mechanical design engineer developing products from railroad instrumentation to drug delivery devices to diagnostic equipment. Greg also has specific expertise in the world of FEA (Finite Element Analysis) and EHD (electrohydrodynamics). 

ABOUT BEING AN ENGINEER

The Being an Engineer podcast is a repository for industry knowledge and a tool through which engineers learn about and connect with relevant companies, technologies, people resources, and opportunities. We feature successful mechanical engineers and interview engineers who are passionate about their work and who made a great impact on the engineering community.

The Being An Engineer podcast is brought to you by Pipeline Design & Engineering. Pipeline partners with medical & other device engineering teams who need turnkey equipment such as cycle test machines, custom test fixtures, automation equipment, assembly jigs, inspection stations and more. You can find us on the web at www.teampipeline.us

***

Valued listener, we need your help getting to 100 podcast reviews. Win a $50 Amazon Gift card if you leave us a review on the Apple Podcasts. Simply email a screenshot of your 5-star review to Podcast@teampipeline.us , the email will be in the show notes. We will announce 5 lucky winners at the end of the first quarter in 2022.

About Being An Engineer

The Being An Engineer podcast is a repository for industry knowledge and a tool through which engineers learn about and connect with relevant companies, technologies, people resources, and opportunities. We feature successful mechanical engineers and interview engineers who are passionate about their work and who made a great impact on the engineering community.

The Being An Engineer podcast is brought to you by Pipeline Design & Engineering. Pipeline partners with medical & other device engineering teams who need turnkey equipment such as cycle test machines, custom test fixtures, automation equipment, assembly jigs, inspection stations and more. You can find us on the web at www.teampipeline.us

Transcript

Presenter: 0:00

The Being An Engineer Podcast is a repository for industry knowledge and a tool through which engineers learn about and connect with relevant companies, technologies, people, resources and opportunities. Enjoy the show.

Greg Trees: 0:16

Just keep working. The problem is what I found is when you have a dedicated high performing team, with, I'll say the right people on the bus, in the right seats of the bus, you can bring home very challenging problems.

Aaron Moncur: 0:44

Hello, and welcome to another episode of the Being An Engineer Podcast. Our guest today is Greg Trees. Greg has a degree in mechanical engineering and has spent nearly 30 years as a mechanical design engineer developing products from railroad instrumentation to drug delivery devices to diagnostic equipment. Greg also has specific expertise in the world of FEA or finite element analysis, which we'll definitely dive into during this interview. Greg, thank you so much for joining us today.

Greg Trees: 1:15

Thank you, Aaron. Glad to be here.

Aaron Moncur: 1:17

So what made you decide to become an engineer?

Greg Trees: 1:21

That's a great question. 'Been asked that many times over for my career. So back when I was, say, my earliest Christmases that I can recall, maybe six years old or so, my dad was fairly handy and he used to give me tools for Christmas from reviewing a hammer or saw, drill, jigsaw, or this kind of thing. So pretty young, my life, I started also taking things apart, fixing my bicycle, or my friend's bicycles and helping fix the flat tire and help with those kind of things. So that was just kind of the start. Then over time, as progresses, you know, again, the school and mathematics you start learning, liking math, Krishna searching, people say, 'Hey, you know, you should be an engineer, you're really good at Math and Science.' So, on top of that, my grandfather, who was one of my role models, I looked up to his background was West Point, and electrical engineering insurance. So I had a lot of respect for him. And he was an engineer. And of course, I didn't really have that electrical, I'll say, flair for things but mechanical savage to me. So my Knossos, the discipline of mechanical engineering, it just kind of sounded like a good fit, although I had no real idea. But everything goes through does for living other than, you know, the design of Carson. So I kind of really kind of just my progression towards moving that direction for as a college major.

Aaron Moncur: 2:40

That's so cool that your dad gave you tools for Christmas. And that makes me think that I should be given tools to my kids. It's coming up here pretty soon. And one in particular, I think would love his own set of tools. What a wonderful idea. Do you think that your your father gifted these tools to you? Because that was kind of the direction he wanted to steer you? Or was it more you'd already kind of shown like this mechanical aptitude and he thought out tools, that'd be a perfect way to help him along?

Greg Trees: 3:13

Yeah, I'm not sure if it was intentional, or he just wanted to see his son use these things. And we haven't really talked about the good for me to follow up and ask him, why did you give me these tools as a young boy?

Aaron Moncur: 3:25

Free labor, right?

Greg Trees: 3:27

Exactly.

Aaron Moncur: 3:27

As soon as you get trained. Alright, well, let's talk about FEA a little bit, it's not something that we have spent much time talking about here on the podcast and maybe share really briefly, what is FEA? What what does that acronym mean, finite and element analysis, what does that even mean? And then a brief summary of how it's used, generally speaking, in a product development environment?

Greg Trees: 3:53

Sure. So FEA stands for finite element analysis. And basically, that's where you take a an object and you break it up into small bits or chunks called elements. And from there, you can assign a material properties to it listen to your properties of stainless steel or plastic or rubber. And you can put forces on loads, you can vet it, you can twist it, you can then have even more complex analyses, where you have a whole assembly such as even automobile, automobiles are virtually crash tested in a computer model so that you have finite element analysis is basically building a computer model to simulate your object with its intended requirements and to try and determine his performance whether or not those comprise a fitter versus building an empirical solid, physical thing.

Aaron Moncur: 4:42

I did not know that entire automobiles were tested in FEA. That's really impressive, in part because I've done a little bit of FEA, I'm nowhere near an expert at all, but I've done a little bit. And whenever I run an FEA, I was trying to strip out as much many parts as I can because I, I'm running it on a computer not not a supercomputer, or a cloud bank or something like that. And it's really processor intensive. What, what does that system look like that runs an entire automobile is this like, you know, Amazon cloud services are something?

Greg Trees: 5:19

Right, we actually have access, where I work to Amazon cloud services. And we also have our own high performance compute cluster, which is actually faster than the cloud. But the cloud has, I'll say more, I'll say expandable hardware so that if you need to solve a larger analysis, which is maybe more than what you have available, let's say I think our current high fours cluster, we have a workers near 256 cores, so to speak, you know, of course, one CPU could happen before a cost. So this allows you to scale up to solve large problems. It's amazing that automobiles, you said are also crash dummies, and they're virtually in airbags deploying herself there. Now, I'm not in the automotive business, I'm in the medical device business. So we're simulating devices and our materials failure. We'll be doing I'll say electro electro structural, thermal coupled analyses, with other types of analyses where you're structuring, or coupling fluids and structures together. And so it's really amazing just the the fidelity of these models and the physics you can coupled together to have a real life simulation. And then the hardware we have today available to us, I'll say in this age, it's just it's just amazing. What we could say no, just because the physical hardware exists in the software.

Aaron Moncur: 6:34

That's incredible. I had a boss, I don't know, 15-20 years ago, who, and I'm sure FEA has come a long ways since then. And it's also entirely possible that he was just ignorant to what FEA could do at that point. But I remember him saying, when when someone brought up, 'Oh, let's do some analysis on this, let's throw it an FEA and see what we get.' He would say, 'Well, we're just gonna have to test it in the real world anyway. So let's forget about the FAA and just go straight to real world testing.' What, what response or feedback might you give to someone with that mindset?

Greg Trees: 7:10

Well, ruble testing is, of course, always what I'll say some of the ultimate. But one things we try and encourage also engineers to do in their development of the design is to develop a good mathematical model. And it couldn't be something to develop equations of Excel or Mathcad, or some other, you know, numerical analysis tool, or I'll say an FEA program analysis, and then you have your empirical tests, what you want to do is you want to compare your empirical results. So say, for a physically built test to your analysis, if this is something important that you're also not just a king, every beam equation like California is going to deflect so to speak. But that's like a really important requirement, you have both your empirical results and your analysis model. And if the analysis and the model agreed, when I'll say a reasonable margin of error, what that tells you is you've minimized the risk of not understanding your design. Now, inevitably, you also you can have disagreement, you can have an analysis which to say, oh, man, the brake force should be 100 pounds, but in real life, if only 50 pounds? Well, that's a that's a pretty large percentage difference. And so what you don't know, just because there's a difference, a lot of times people gravitate, well, the empirical, my physical test must be correct. Well, in truth, you don't know because oftentimes, many types of tests are actually improperly set up. And the results can be poor. So you might actually find that right compared to your analysis of you check your boundary conditions of your physical test find maybe that this causes, or no, it's maybe the material properties, or the boundary versus my FEA model were inaccurate. So the the idea is to try and to have both, so that you can get confidence that you haven't missed something big.

Aaron Moncur: 8:57

That's a really interesting way to think about it. I always thought of FEA is like a first pass. Maybe you try that just to make sure you're in the ballpark, and then you do physical testing, I hadn't thought about it from the standpoint of the FDA could actually be a check against the the the physical testing to tell you maybe your physical testing isn't right. That's a really interesting point. Can you talk a little bit about the differences between linear and non linear FEA and when to use each?

Greg Trees: 9:26

Oh, well, so a lot of simple models are in the linear region of analysis. So let's say you're doing bending of the kingdom for beam, and you put a load on the end of the beam, and on the other side of the beam is fixed. If you put the load on the end of the beam, and deflects let's say, a 10th of an inch was applied your route the load and load comes all the way back. That's an example where he did not take the material beyond I'll say a linear condition. So for metals, let's say the linear elastic point would be the yield point. So or a stainless steel, it might yield, I'd say, it's not very well he treated maybe 36,000 PSI. Beyond that he might need a nonlinear material model to capture, I'll say, the material beyond I'll say, the the point of yield, which is also in the plasticity region of material. And those are more complex material models, which we do all the time. But you need to recognize that anytime your stresses in your model are beyond also the yield point, your if your material properties are only in the linear region, you're no longer necessarily predicting things accurately, you might have a small margin of error, you might have a large margin of error. Now, that's simply non linearity in the area of material properties, you can also have contact between components. So if you have ball bearings, you know, rolling on a race, so to speak, contact is highly nonlinear with regard to saga net model to solve and getting good convergence. So you can have material non linearity, you can have really cheap metric non linearity. Try to pick sources. Here, I think there's other sorts of nonlinear work, which is a couple of common examples.

Aaron Moncur: 11:08

So if you're doing an analysis on a metal material, it's pretty clear where that linear range is, if you look at like a force versus displacement curve or stress strain curve, if you're looking at a plastic material, that that linear range is a little bit less clearly defined. Our plastics, another area where nonlinear analyses are performed?

Greg Trees: 11:33

Oh, yes. And plastics can be highly nonlinear. I mean, kind of by definition, when we say the straight part of that Young's module, so to speak, we talked about, it's somewhat straight, anaplastic, but very quickly, or fairly early on, it starts to get curvature to it. So often, in a plastic, if you're doing an elastic plastic model, which is what we call a binding or material model is basically using two lines to try and simulate that material that's relatively simplistic, oftentimes good enough to capture a certain amount of I'll say, fidelity in the model, if you need more accuracy, having a damaged model where if your plastic strain goes down a certain level, certainly those models become more sophisticated to capture higher levels of strain. And to properly capture, I'll say, your histories as well, if you hold the material so far, it comes back a certain way when the loads released. So plastics are can have very sophisticated, I'll say, properties interacting with our office,

Aaron Moncur: 12:35

How much experience or material science knowledge, or just material properties knowledge, does one need to have to run FEA effectively? Is this something that if you're really focused on it, you could pick up in a few months? Or is it really years before, you're really very competent with it?

Greg Trees: 12:56

Let's say what's really cool about the time we live in is a lot of the most simplistic analysis, people start out with our single component analysis. And they're just trying to understand, you know, is my wall thickness thick enough? Is the part just gonna be strong enough? In my beyond yield? They're trying to ask basic questions for the basic answer. So the I'll say some material cracking, FEA packages have default materials already kind of loaded. But there's simplistic material models, often just linear material models, and they don't capture I'll say, nonlinear behavior, such as plasticity. So if you're doing the cantilever beam, and you find in you load it up, if I all my stresses are below the yield point, hey, this model is great, it's perfect, I can make my decisions I move on. If you start to see that your stresses are much beyond they'll say that the point like, well, I might even invest in a nonlinear material model. But with regard to picking it up, I'd say even the cad tools we have today, which have vetted FEA like pro engineer SolidWorks, I think almost have some sort of vetted FEA at this point. The tools are very simple to use. And I think people can make reasonable I'll say, without much say, guidance, reasonable material, I'll say boundary conditions and settings, which allow you to make good decisions. But in general, though, winning with latte learning offset tool like this, there's an old phrase right in cheering garbage in equals garbage out. If you have a poor boundary conditions are the most notorious way to have had a VA results if you artificially constrained something, which is called prejudicing it. If you put too large of an error, you say this is just fixed, it's not movable, what you've just done is you've made it infinitely rigid. And so when you put your load on it, you're fine. Your stresses are very low, but that was an artifact of your boundary conditions. So it's always good when learning a new tool like this to have I'll say someone who's more senior look over your results to give you some tutoring as far as saying you know this boundary condition primacy adjusted to be more realistic so that you're not having non conservative results. Often we talk about having results which are more on the conservative side, like over predicting stress versus underfitting under since.

Aaron Moncur: 15:00

That does make sense. Can you think of an example that you can share where you and your team used FEA as part of the product development process, and it was really helpful it you look back and think to yourself, 'Thank goodness, we did that, it saved a lot of time or a lot of money,' anything like that.

Greg Trees: 15:19

There's, I'm saying, for us, that's all the time. FEA is a kind of necessity term trying to avoid I'll say do moves, especially in the medical device world, we're often if you're putting devices down to five millimeter trocar, around Egypt, very small parts, with a very high lows, you're usually stressing the price really high. So we're really trying to, we're really pushing the materials kind of to the maximum, so to speak. So oftentimes, we'll have to do things such as this if we have five components in an assembly. And if I have a imagine, if you will, a pair of pliers, hire apparel, hires eventually kind of come down to roughly a parallel gap towards the end of your squeezing something. Now imagine if you want that gap to be the width of one human hair, and it's a one inch long gap? Well, if you have four components, which make up your pliers, and you're trying to achieve that uniform gap, say between two and four thousandths of an inch across that, that range. What we had to do in order to I'll say design that is doing what we call a virtual design of experiments, where we'll do several, I'll say test cases, let's say 16, unique FEA models, where we will have the device in various configurations and the components of various sizes and dimensions to try and understand what particular attributes to the design are driving the effect to get a uniform drug out. And from that, I'll say analysis, we can develop a transfer function and visually help us I'll say, finalize the design effort tolerances on the key dimensions which control the final output of design and manufacturing process.

Aaron Moncur: 17:02

Interesting. Yeah. Are there times where we shouldn't be using FEA, use cases in which it's just not appropriate to use FEA?

Greg Trees: 17:13

So as a pragmatist, you know, I try and do whatever makes the most sense for a given case. In other words, you try not to do through FEA everything per se, there's oftentimes people want to use FEA for something where the physics just really hasn't been developed yet. That technology is not there yet. You try to think of just a new sample here from...

Aaron Moncur: 17:35

RF, maybe?

Greg Trees: 17:36

Oh yeah, let's say let's say the physics environment volved with just sealing a vessel, within a medical device, there's, there's a lot of physics going on, you have electricity flow, you have joule heating, you have mechanical bending, you have the phase change going on with the vessel as it's sealed to the heating process. That is, obviously to do all that numerically, which would be wonderful as a great aspirational goal. I don't know of any policy accompanying which has also cracked that just yet. So sometimes you're left with doing things empirically. Because there's the technology is not there, you can cry. And you can do pieces of parts of it numerically, but not the whole thing. Similarly, sometimes it's just if you have a bunch of parts sitting over on the bench, and you have a load frame where you can just go walk over in full test them, versus also developing a real sophisticated VA model, which might take you you know, if it's really sophisticated, you have several days to develop and a couple of days to run, it might be quicker to test. So you can do whatever, whatever gives you the reasonable information, the quickest speed that you could trust to make your decision.

Aaron Moncur: 18:43

Yeah, great advice. How about for, for those of us who are just getting started with FEA, where we're very junior, we're still learning what the tool is and how to use it, or are there any areas that you can recommend we really watch out for, what are the gotchas that can really get you into trouble if if you if you don't even know enough to be aware, to watch out for those areas?

Greg Trees: 19:11

Well, in FEA, the two biggest ones are your boundary conditions and the loads and how you apply them. So one of the first mistakes people make is they'll, if they if you have a pope let's say you have a part of flowerpot sitting on top of the table top desk and the flowerpot has it's full of dirt so to speak on the inside or fill out this paper full of water makes a little easier because water applies a pressure on the inside of the flowerpot. Of course that firefighters sitting being supported by the table and so on my initially said you know what to do with the bottom that flowerpot I'm going to fix that if that rigid boundary condition. And what it's just done is by fixing it make it immovable. He also made it infinitely rigid. So zero stress can develop any note nodes any elements that are linked to that boundary condition. So now there, the stresses will be artificially low. And it may lead you to conclude that your design is strong enough, but it's really not. The next thing would be how you apply your loads and low. Sometimes might people might be tempted to do a point load, like all the force and tickle like a single dot, a node was what we call that, verse or mate put on a line edge. And what those kind of inputs will do is the artificially tried to stress up too high, which thinking leave and conclude that, you know, there might be an issue of design, really, it's just how the rules are applied. And this concentrated stresses just make your stresses look high or higher than they really are in reality. So just be careful with your boundary conditions and higher lows.

Aaron Moncur: 20:44

This really gets down into the weeds that I don't know how effectively we'll be able to discuss it just using verbal communication or oral. But let's say that you have a device a medical device a handle, and you want to understand how hard can a person squeeze on this handle before it breaks or exceeds the yield point plastically deforms. An easy way to apply a load to a handle like that would just be to isolate a defined surface. Maybe it's kind of like a oval shaped surface on the face of this handle and apply your load across that surface. You have a load applied distributed evenly, evenly across this oval, you can see the oval represents the person's thumb or something like that. But But even when you do that, the load is at its at its its fully distributed maximum value up to the edge of where you have defined that oval to stop. And then, a thousandth of an inch beyond the surface defined by that oval, there's no load, which is not really how things work, right? If someone's thumb is pressed up against a material, there's going to be a gradual reduction of load. As you move away from the surface under that thumb, is there a way to accommodate for things like that? Or is that just a limitation right now, of FEA?

Greg Trees: 22:11

There certainly is a way, I mean, if one was really concerned about the stresses directly beneath the finger and the squishiness of our flesh, so to speak, certainly you could develop an assembly model, we had contact and you have all say, an appropriate squishy material representing the flesh or muscle of one's finger

Aaron Moncur: 22:28

Got it

Greg Trees: 22:28

So sheen on that, that could be done that level of fidelity. We do those types of exact scenarios you describe all the time with regard to concern about I'll say the physical force of some concern on the handle and breaking it. And oftentimes, the peak stresses are not right below worsens, applying a load that picks stresses are often attending, right. So venting is usually a distance away from where the load is actually applied. So it's often not very crucial to I'll say, model on that fidelity, right? Anything where the lives of clients and just making sure that you've got the right four separate location. Getting right at the right moment, I'll say the venue stresses for somewhere else in the park, which is huge. So if you would have I'll say, a point load or line load, or having contact to a rigid cylinder just at the right distance, you would get good results in the area that care.

Aaron Moncur: 23:20

That makes a lot of sense. Great way of explaining it. Alright, I'm gonna take a very short break here and share with the listeners that teampipeline.us is where you can learn more about how we help medical device and other product engineering or manufacturing teams develop turnkey equipment, custom fixtures and automated machines to characterize inspect, assemble, manufacture and perform verification testing on your devices. We're speaking with Greg Trees today. Greg, you you worked at a company called Battelle, which we've actually had other people from Battelle on the show several times in the past, actually. Anyway, at Battelle, you developed IP in the area of electro hydrodynamics or EHD, can you share a little bit? This is not an area that I'm familiar with. So I'm guessing there are a lot of engineers out there who are not very familiar with this area, what is EHD? And can you share a brief example and application of how it's used?

Greg Trees: 24:17

Sure. So there's a lot of different types of, I'll say lesser heard of types of physics, and EHD, electro hydrodynamics is one of them. So basically, if you have a I'll say, a flat plate, a ground plate, your table that you're sitting at right now is ground, if you held your pen, your ink pen above the table a couple inches, if you had the pen I'll say the voltage of 10,000 volts and your ground plate as this is zero observables ground. If you have the right fluid properties, the right voltage potential you can get that fluid to analyze into an aerosol. So instead of your fluid and come to also write it on the pen on paper, so to speak. If you turn your pen, it'll say a sprayer, so to speak, so that fluid will form what's called a Taylor cone. And that that take the tip of that little cone of fluid is very high electric field source. And that the height, the charge charge density, and the gradient overwhelms the surface tension of the fluid in positive bursts into an aerosol. And that aerosol is charged as a very high amount of charge per unit volume on it. So then the particles rapidly repel from each other. So becomes one of the interesting properties of HD process fit so, if tuned properly, is just like a monodisperse aerosol, meaning the distribution of the particles coming from that process are very small, their size, so you're targeting at the time for drug delivery in the lungs to have particles between maybe two and 10 microns, something of that nature. And this process can help a very fine distribution of particles. But what's also interesting, at least electro hydrodynamics, initially, that aerosols charged, so all the particles are rapidly I'll say repelling each other. And of course, if they get near a source, like your body, or paper or a metal plate, they want to plate it. Versus if to get lungs, you then need to discharge the particles. So we also had to combine that with a methodology to fire much of ionized air added all those little particles to neutralize their charge. So that could be I'll say, inhaled into the lung and get get very good. I'll say drug deposition. Now they're things

Aaron Moncur: 26:33

Interesting, that that makes me think, I live here in Arizona, where it's super hot. I wonder if there's an application for that in misting systems might be maybe too expensive of a process for that. But she could get some small particles, I think the diameter of the water is is directly related to the cooling effect it has. That would be interesting. Yeah. Anyway,

Greg Trees: 26:58

It evaporates very quickly. So speaking, also, you cooled on the air for you not to worry about getting it straight, although you're getting sweaty. That's a cool thing.

Aaron Moncur: 27:08

Yes, yes, absolutely. All right. Well, let's see what, what is one of the most challenging product development projects that you've worked on in your career? And what were one or two key takeaways from that project?

Greg Trees: 27:22

Well, I say developing vessel sealers is extremely challenging, that we do right now. Obviously, finding a method to deliver energy in a very precise manner, to understand the tissue properties, and all the variables of the human device control, those little indictment for type framers, to get a very repeatable outcome across a variety of tissues is extremely challenging. So basically, one its ethical basis, a bunch of vessel sealers, so to speak. And cutting devices, to that is an extremely challenging technology. And one of the challenging things about it is, since you cannot I'll say this presently fully developed a numerical model and computer model to help you do this, and you're doing a lot of empirical work. So that just puts an extra strain or diligence on the quality of that work insurance thorough who just needs a lot of work.

Aaron Moncur: 28:25

Sure, yeah. I'm curious, what would what would the Greg have today, say to the Greg of 30 years ago, back when you were just starting out as an engineer that you wish you had known back then?

Greg Trees: 28:38

I definitely got some points there. So I think for engineers starting out, I think focusing on your technical for the first time, 5 to 10 years of your career is a very wise thing to do, making sure you're just growing a lot of areas as much as you can trying to get if you're interested in being isolated. And as you're doing technical things, make sure you're taking looking for opportunities to really just develop your technical. After that, I would say learning structure and solving farm techniques. Oftentimes, when you come out of college we know areas of physics and mathematics and science. But you have been really taught sound methodologies for solving problems such as Six Sigma methodologies such as to make and demand the learning about those methodologies can really help you solve problems more efficiently and correctly versus thinking you've solved it, but really, you have some fantastic. The other thing, I'll say, is learning the value of risk management tools and system engineering tools, as you select your concepts as well as you refine your designs. You, I think designers or engineers in general don't recognize that really, a lot of their engineering careers are about managing risk and trying to avoid things from going wrong. So really, a large part of our lives is risk management or really risk managers, which managers have a source for, we just don't realize it. So you're trying to keep something from failing for working. And the antithesis of that, is it not working. So you're trying to constantly balance risk, so to speak. And lastly, I'd say, be a jack of all trades and a master of one. Some people will saying,'jack of all trades, master of none.' But I encourage people you want to really go after something and go after it deeply, is what will happen is if you become a master of something, you'll eventually build a augment or bolt on other things to it. So for example, many years ago, I started going deep into FEA, which then pulled me into knowledge of materials, which then pulled me into knowledge of crazing and other failure modes for material properties. And so allows you to build out a platform if you go deep on something. So mass of all trades, checkable. masses, or none. You got it.

Aaron Moncur: 30:54

I got it. You've you've mentored, I think quite a few engineers in your time. And I was wondering, what are some of the topics that you commonly hit on? It might be similar to some of the things he just shared about, what does Greg today share with Greg 30 years ago, but in your mentoring sessions? What are some some topics that come up commonly?

Greg Trees: 31:16

Well, I think we definitely talked about the technical, making sure people are doing structured problem solving, making sure they're managing the risks, we definitely talked about the balance between going fast. And going through, there's a whole saying, sometimes around, have sometimes going slower is going faster, is sometimes you know, we think a little bit fast, but we could be headed for a cliff, so to speak Korean, basically a novel solution. And then you really haven't moved quickly, you just spent a lot of time and dollars for progress. And we'll get to the objective. common things that people also worry about our statistics these days, I remember, let's say 25 years ago, of course, statistics was important to engineering, also much more. So now, being understanding your design, as far as if you're making a lot of something, you know, hundreds of 1000s of something. But now what starts mattering is okay, how good is it for example, on the transistor, microprocessor yet billions of transistors, right? So if you have one out of a million transistors is not functioning, that's a large number of transistors that aren't functioning on the ship. So they are obviously used to that kind of methodology, and how to deal with I'll say, the non functioning transitions of ship, if you're developing, that's just on one product. Yeah, that billions of something, we typically I'll say, think about cars, you know, 4 makes 600,000, you know, F150s a year. So that's, if only one out of 600,000 had an issue, or that CEO, two out of a million. So worked on jobs to try and do with your testing, what is the reliability or the capability of your design, with a car to 4 million units, so to speak, versus just showing that you can make one work, you really need to demonstrate that millions work, so to speak.

Aaron Moncur: 33:08

Yeah. Can you share maybe one or two of your favorite vendors? I'm talking about people like, or organizations like McMaster Carr, or I don't know, maybe there's some kind of like 3D printing service bureau that that you really like. Organizations like that or or vendors like that, that listeners might find useful in their their own projects?

Greg Trees: 33:32

Oh, yeah. Well, certainly McMaster Carr's probably never engineers toolbox per se. I think for prototyping was really interesting with come on the senior the last several years, I've seen proto labs, initially, having prototypes made a lot of smaller machine shops, but how they automated the process of getting parts and the number, decent quantity and the quality is changing, but for labs have done for people as far as 3d printing and prototyping, and just how they all say commoditize the ability to get prototypes quickly. Another really interesting thing, as far as it's kind of come about, so to speak.

Aaron Moncur: 34:06

Yeah, yeah, we use them regularly, if not often. What are a few habits that you've developed that have proven useful to you, it could be engineering related or personal.

Greg Trees: 34:21

I'd say continuous learning will help you they say engineers, from the time they graduate to over the course of the career will need to kind of relearn their craft, you know, two or three times. So it's definitely good to try and stay abreast of what's going on, because things are obviously constantly changing. So you talked about FEA today. You know, when I graduate from school, of course, FEA was available when I was in college, and I looked into that, but now I say the 1 level's continuously rising so to speak, the capabilities are just constantly being increased. So in other words, FEA, I think it's now for the superior mechanical engineer as the board is kind of a table stakes you should be able to have Is this kind of like a Swiss army knife need to have that in your bag, so to speak? The question is, to what level do you want to, I'll say, take that in your own skills. We've talked a little bit about structured problem solving. Some of the habits, I'll say, just hobbies, I'll say, in general, like I enjoy working with cars. And partly just because I have an old fleet, so to speak, and to take care of, you know, to learn all the new cars, trees, houses. So I think as a chance to work on troubleshooting things and seeing things I'll say that other people have designed and where they failed, and just practice troubleshooting skills. So being a good design engineer nice, you, of course, need eventually know a lot of processes. I think it helps to familiarize yourself with stamping, injection molding, casting. And the more things you know, in that regard, the more helps you see other ways of doing things, right, just the broader your horizons get. So continuously learning about all these things just really helps you do better at also engineering in general.

Aaron Moncur: 36:01

What's one of the best parts of your job, and also one of the least enjoyable parts of your job?

Greg Trees: 36:11

'Course in the product development continuum, right, you kind of start off with I'll say, the Fuzzy Front End, we have an idea and a concept. For me, that's kind of the fun part where it's like, okay, you're really just trying to sculpt you were the customers and the needs, were the potential solutions, start identifying concepts and meet those solutions. That's the real fun part where it's, it's like fast and fluid, and then have to kind of narrow things down to a final concept. And you got your, say your project charter, business commitments when you're on delivers to the business. Now things get a little more rigid and locked down. They're still fun in there. But I'll say that's when also that paperwork, at least in the medical device industry starts to increase quite a bit. So you'll be doing engineering bills, as I verification testing, lots of documents circuitry for you. And so you kind of looking through this process multiple times as you get to the end. So for me, of course, I'm actually I don't know too many people who really get engineers to get jazzed up over doing a lot of documentation. Although it is extremely valuable. In the medical device space, it has a real sincere, honest, good purpose. I'm just call it the fun part of engineering, but it's a very necessary part of it.

Aaron Moncur: 37:22

Yeah, yeah. It's, it's always a little disheartening. Not necessarily to me, because I've seen it so many times. But I've heard this sentiment from our customers who themselves are not engineers, where we'll, we'll start on a project. And it's that beginning, like you talked about, right, the kind of fast fluid part where you're, you're architecting, something you haven't really worked out the details yet. But you're architecting something. And in a surprisingly short amount of time, you can come up with a design that looks pretty good, you know, someone who's not an engineer might look at it and say, Oh, we're 80% Done. And we've only spent 20% of the budget, we're, we're in good shape here. And then it can be a little awkward when we're down the road a few months later, and the customers like, 'I saw, you know, months ago that we were almost you're 80% done. And here we are, and we're we're still working on this.' Well, the reality is that 80% of the perceived work gets done that first 20% of the time, but really, the last 80% of the time spent is where all the you know, the details get ironed out. And that can be hard for someone who's not an engineer themselves to to understand

Greg Trees: 38:36

Very true, especially the quality quality of prototypes these days, they look so good. They'll say upper management or so as a marking season. 'Wow, looks like you're done.' We knew it's gonna take you two more years to get this done. All right now. It just doesn't hold the eyes dotted, the T's crossed and all those other things which go along with it.

Aaron Moncur: 38:54

Yeah, you might be selling some lawsuits at the same time. Release this right now. All right. Well, let's see. I've just got one one more question for you. What is his pretty open ended question, feel free to take this in whatever direction you think would be most useful to people listening? What is one of the greatest things that you have ever learned?

Greg Trees: 39:13

Well, I will say I'll try I will delve deeply this but I but I enjoy the topic of theology. And that's a great thing, which gives me peace and the rest of my life. I will go further on that area because the place to do it. But I think from my from a team and engineering team standpoint, what I'll say is, over the course of many years, in many projects, we are faced with these challenges, which sometimes seem sometimes they might be insurmountable, but the vast majority of times they're not. And when the little phrases I want to hear or passages to say was just keep working. The problem is what I found is when you have a dedicated high performing team, with a I'll say the right people on the bus, in the right seats in the bus, you can bring home very challenging problems. And you, it's that confidence that people get after I'll say, living through a lot of very difficult times. If the cartoon just facing tough challenges where the team maybe has a setback, things didn't go right. And you continue to keep your nose to the grindstone and the team stays positive. And you get to the other side, and you finally have enough of those experiences that you're not. Obviously with the old, this isn't rocket science, so to speak. But obviously, because that's also a different level of aerospace is much more challenging level of interiors is all the more cars, jewelry, that kind of work. But obviously, those people get it done. And look, we have aerospace, we have airplanes, we have rocket, so to speak. So most problems I'll say can be solved unless so is asking something which is just not moving around with physics, or within a budget. So my guess my advice to folks would be, you know, you got the right people. Or if you don't think you're at the right town to sell from find right talent, because the problem more than likely is solvable.

Aaron Moncur: 41:08

That is fantastic. That's so encouraging and motivating. We're working on a pretty tough problem right now. So hearing words of wisdom like that, to me personally, that's, that's very helpful. And I'm sure that there are a lot of other people out there listening, for whom that is just as helpful. Well, Greg, thank you so much. I really, really appreciate you taking time out of your busy schedule to talk with me and share some of your wisdom and insights. Is there anything else that we haven't talked about that you think we should talk about?

Greg Trees: 41:42

What they will say too, is just the failure. You know, while often we try and avoid failure in our career as much as possible in our lives, but it's kind of inevitable in the dream world that you will experience setbacks and failures. And I and I have had, I'll say, experience some of those, and they are rough at the time. And it's not also fun to go through those experiences. But you will learn from those experiences. And those will obviously help you teach others about where those pitfalls and roadblocks are. I think there's some of those lessons, which are hard learned. I think it's a matter of not how you fall down. But it's how you pick it back up from those experiences, even at not as just as an individual or as a team. So just remember, usually you're in it with a team. And I think it's how you lead and your how your the team handles that scenario. We'll turn the roll, how they eventually that that effort pans out how they approach it, to let it knock it down and how they get from that.

Aaron Moncur: 42:44

Amen. Amen. Oh, and I wanted to touch on your comment about theology very briefly. It's, it's super interesting for me to hear people talk about it because I myself am a a man of faith. And I have never tried to facilitate that kind of conversation here on the podcast, but it comes up fairly regularly. You know, people just start talking about it. And it's just interesting to me. How organically that that comes up. And anyway, I think it's neat. So thank you for, for sharing that perspective, as well.

Greg Trees: 43:23

Thanks, Aaron.

Aaron Moncur: 43:24

Well, how can how can people get in touch with you, Greg?

Greg Trees: 43:29

I think LinkedIn probably one of the easiest ways to get in touch with me, that's out there.

Aaron Moncur: 43:35

Excellent. Okay. Well, again, thank you so much. This has been just fantastic. And I can't wait for for all of the listeners to hear everything that we talked about today. Thank you, Greg.

Greg Trees: 43:46

Thanks, Aaron, I appreciate it.

Aaron Moncur: 43:51

I'm Aaron Moncur, founder of Pipeline Design, and Engineering. If you liked what you heard today, please share the episode. To learn how your team can leverage our team's expertise developing turnkey equipment, custom fixtures and automated machines and with product design, visit us at teampipeline.us. Thanks for listening.