Getting geeky

"luvMeSome142 & some Lincoln!" (luvmesome142)
10/06/2015 at 14:25 • Filed to: None

The picture will be explained shortly.

So, I just took the Town Car in for an inspection and I need a new upper ball joint. I forgot to ask about a pickle fork at Autozone, so I did a quick search on how to remove a ball joint without one. I’ve done it before by threading the nut down to the end of the bolt and beating on it with a hammer, but I thought there might be a better way. Turns out there is!

I found !!!error: Indecipherable SUB-paragraph formatting!!! of hitting both sides of the taper joint with two hammers at the same time. They described the action thusly:

hammer either side of the taper at a slight angle. This squeezes the bottom of the housing that the taper fits into, and opens up the top.

I thought that was an interesting description, but what does that actually look like?

Being an engineer and having access to finite element analysis software, I did a quick study of it.

I made a very crude approximation of the suspension bit with the tapered hole in it. Then I applied an arbitrary force to a few nodes where I imagine the hammers would contact. About 1000 lbs per side. What I was looking for was the deflected shape, not the stresses, so the force wasn’t important.

The results clearly show that the bottom hole is squeezed and the top doesn’t deflect much.

It looks like I picked realistic forces after all, because the points where I applied the load are just about at the yield point for steel. As you might experience, when you hit a heavy piece of steel with a hammer, there is a slight flat spot or divot. The metal yielded. You can also see that the stresses drop way off away from the hammer blow. We are no where near breaking the part with this load.

I’ll tell you how easy the ball joints come out tomorrow!

DISCUSSION (36)

Sweet Trav > luvMeSome142 & some Lincoln!
10/06/2015 at 14:32

two prong pullers can work great too!

luvMeSome142 & some Lincoln! > Sweet Trav
10/06/2015 at 14:35

When you have two big hammers, everything looks like a steering knuckle.

RamblinRover Luxury-Yacht > luvMeSome142 & some Lincoln!
10/06/2015 at 14:36

This is, realistically, something I do nearly every time I remove a tie rod end. Usually a single hammer, though, sometimes with a backstop, sometimes without. It also helps to have a pickle fork tapped in or something to provide tension.

This is my jam, so to speak, because I have rarely replaced a tie rod end that was *not* over thirty years old. The ball joints on my ‘63 Ranchero required this *and* pickle fork and repeating the process until my hands were sore - they were just that sticky. No fire needed, though!

64Mali > luvMeSome142 & some Lincoln!
10/06/2015 at 14:42

I both applaud you for checking to see if this works and hate your for giving me nightmares. I did FEA for 3 years, it can get boring fast. What program are you using ?

RamblinRover Luxury-Yacht > 64Mali
10/06/2015 at 14:51

He might be using Siemens NX. It’s what the NASTRAN suite in NX looks like from my memory, though it’s been a while.

Seat Safety Switch > luvMeSome142 & some Lincoln!
10/06/2015 at 14:52

Finally, a use for an engineering degree.

luvMeSome142 & some Lincoln! > 64Mali
10/06/2015 at 14:54

It’s been about 3 years for me now. We’ve always got something new to work on, so it’s still interesting to me.

I’m using Nei Nastran with FEMAP mostly. Occasionally I’ll jump over to the machine with ANSYS, but only when the customer specifies it.

ly2v8-Brian > luvMeSome142 & some Lincoln!
10/06/2015 at 14:55

When in doubt hammer it out.

luvMeSome142 & some Lincoln! > RamblinRover Luxury-Yacht
10/06/2015 at 14:58

The car is only a ‘99 and pretty clean underneath. So, hopefully, it will pop right out. But, you never know.

DasWauto > luvMeSome142 & some Lincoln!
10/06/2015 at 15:12

Hah, only an engineer would do this. Now I kinda want to acquire some FEA software.

Bandit > luvMeSome142 & some Lincoln!
10/06/2015 at 15:15

Kinja'd!!!

Dat FEA. Ansys? Or what? I’m not familiar with the different brands of software?

luvMeSome142 & some Lincoln! > Bandit
10/06/2015 at 15:20

FEMAP with NeiNastran solver

Xyl0c41n3 > luvMeSome142 & some Lincoln!
10/06/2015 at 15:22

You need to make more posts, please. :)

luvMeSome142 & some Lincoln! > DasWauto
10/06/2015 at 15:27

I know, right?

There are several open source packages out there, but the learning curve is steep. I downloaded CAELinux a while back, but didn’t have the time to mess with it much.

luvMeSome142 & some Lincoln! > Xyl0c41n3
10/06/2015 at 15:28

Hey, thanks!

uofime-2 > luvMeSome142 & some Lincoln!
10/06/2015 at 15:30

My technique was to support the control arm then bash the knuckle with a large hammer.

Much easier than trying to swing two hammers perpendicular to gravity simultaneously.

DasWauto > luvMeSome142 & some Lincoln!
10/06/2015 at 15:31

I may have a copy of Catia somewhere that has some basic FEA in it. I haven’t played with that for almost 2 years now though. :<

tromoly > luvMeSome142 & some Lincoln!
10/06/2015 at 16:40

What does the deflection look like when you turn scaling off?

It’s been a while since I’ve done FEA work, wouldn’t the taper of the ball joint change how the stresses are distributed because the lower end of the taper can’t collapse inward?

luvMeSome142 & some Lincoln! > tromoly
10/06/2015 at 16:51

You can’t even see it. It pinches in about .0002” with the 1000 lb load.

And, yes, with a pin in there the stresses would be different. I was just doing something simple. You could really turn this into a science project if you wanted to involving nonlinear materials, dynamic loading, friction between the pin and the hole.

Kailand09 > luvMeSome142 & some Lincoln!
10/06/2015 at 19:04

Hey do you do analysis as your focus?

I’m dabbling in FEA at work and the only engineer interested in it.

Anywho really cool use of the tool, thanks for sharing.

luvMeSome142 & some Lincoln! > Kailand09
10/07/2015 at 08:42

Most of my work is analysis and most of that is FEA, but I also do design work, testing, and whatever else comes along. I’m in a two-man organization, the owner and me. We do engineering consulting and service for a handful of clients that are enough to keep us employed.

I wasn’t exposed to FEA before I ended up here. I really took to it. If you are interested in it, keep at it. Let your coworkers know what you can do, and how you can help them.

Also, refresh yourself on stuff you learned back in school, dynamics, statics, mechanics. You want to be able to do quick hand calcs to validate your models. FEA can bite you in the butt if you are not careful. It can be extremely embarrassing if you show your boss some results and he points out that your off by a factor of 1000 because you used the wrong units.

Kailand09 > luvMeSome142 & some Lincoln!
10/07/2015 at 17:20

Very interesting, thanks! I too find it really cool and want to pursue it further. Consulting seems like a really fun gig.

When you started, were you mentored? I have read in some simulation circles that is the best way to learn.

I’m a bit lost on how to start, but I definitely agree in needing to get back on point with dynamics etc. Honestly didn’t get the best education in that regard where I went to school, oddly enough.

Maybe I should check for a low cost version of high end FEA software to test with as well. To be honest, dynamic and non-linear is stuff I rarely touch as of yet. Don’t really know how to carry those out.

luvMeSome142 & some Lincoln! > Kailand09
10/07/2015 at 21:47

I was mentored more in how to think about the analysis than in how to use the FEA software. The software came with a bunch of examples that I worked through and that got me off to a good start. But the fastest learning comes from when you are working on a real job for a client.

If the company you work for doesn’t have a seat of some software, there are some open source codes out there. There are also some cloud based options that are getting pretty capable.

I know NeiNastran can be used for free for very small models (node-limited). The tutorials can all be done under that limit, I think.

luvMeSome142 & some Lincoln! > RamblinRover Luxury-Yacht
10/07/2015 at 21:54

Turns out it was a banjo fitting. I loosened the bolt and it popped free! I didn’t need the hammers at all. I was kinda bummed, really.

Kailand09 > luvMeSome142 & some Lincoln!
10/07/2015 at 22:57

I may try that.

I do have access to a seat of SolidWorks Simulation Premium (full content), but I feel SW is more limited and less utilized in the industry.

The “how to think” part I suppose will come to me with self-learning. There is an MIT opencourseware class or two on it, but I’ll probably have to hit that after I brush up on basic mechanical concepts. Wish I was more into it during college / the program actually utilized the statics/dynamics but that’s a whole different topic...

Kailand09 > luvMeSome142 & some Lincoln!
10/07/2015 at 23:01

Also, any advice on how to get into the consulting world?

I still have learning to do and getting better at design, etc... However, I am constantly trying to reflect on myself and what it is I truly want to do... Closest thing I can come up with is something that involves change. Maybe I’d find that consulting.

RamblinRover Luxury-Yacht > luvMeSome142 & some Lincoln!
10/08/2015 at 07:43

Well, there are enough taper-style ball joints out there that this is useful information - not to mention the application to tie rods. So, not a waste figuring out that this works at all.

luvMeSome142 & some Lincoln! > Kailand09
10/08/2015 at 09:47

Constantly having new projects is definitely a plus for me, too. I love to figure stuff out. That being said, you could find yourself always doing new things in a non-consulting job, too. I’ve changed jobs pretty frequently and always to small (and very small) companies. Most of those companies made a physical product. The staff was always pretty small, and there was no shortage of new things to work on.

This is a bias of mine, but I’ve always felt that if I took a job at a big corporate company, I would get stuck doing the same task over and over again. Out of college I turned down an offer at a giant shipyard where I would be one of 72 mechanical engineers in a giant room of cubicles. I ended up taking a job at a smallish Navy contractor where I was the only degreed ME. I’m not sure which one was the better choice.

As far as getting into consulting, some networking might be in order. If you do some googling or searching on LinkedIn, you should be able to find some people to talk to. It can be a little intimidating, but most people are flattered if you reach out to them and ask them to tell you about their job and their industry. Don’t contact people to look for a job! Just find out what their corner of the engineering world is like. I wish I had done more of that early on.

64Mali > luvMeSome142 & some Lincoln!
10/08/2015 at 21:21

ANSYS is such a horrible program. I’ve had crashes after running an analysis for 32 hours.

Kailand09 > luvMeSome142 & some Lincoln!
10/08/2015 at 21:44

I can see that, true. Where I’m at now is a medium sized company, and privately owned. Unfortunately it seems like some of the directions don’t really align with me, and it is a limited product range and fairly repetitive seeming.

Thanks again for more advice. I will certainly work on my networking. One great thing about my current employer is they are open to conferences and the like, which is also a great place to just learn from people and their experiences, and networking. I need to come out of my shell a bit more with that I think.

luvMeSome142 & some Lincoln! > Kailand09
10/09/2015 at 09:34

I’m quite introverted myself, and I’ve always feared looking stupid. Putting myself out there is hard to do.

One thing that I’ve been doing recently is attending meetups I find on meetup.com. Last month I went to a programmer meetup. I didn’t understand most of what they talked about, but it was a great experience and I met some cool people. If there were any engineering meetups locally, I would definitely be going to those.

Another idea: see if there is a local ASME chapter where you are.

luvMeSome142 & some Lincoln! > 64Mali
10/09/2015 at 09:37

Running long solution time analyses is always scary. I occasionally have an overnight run, and rarely one that I’ll let go over the weekend. The worst is when you go through the results and you realize you screwed up a load or boundary condition.

BTW Femap has more than its share of crashing issues.

64Mali > luvMeSome142 & some Lincoln!
10/09/2015 at 09:48

I had to do some Icing Analysis for an aircraft inlet. We outsourced one of them to the NASA facility that designed the software. We crashed their super computer and actually burned out a few processors after it had been running for a week.

I had to do a lot of thermal and modal analysis which all took 24 hours minimum. Some of the thermal ones took 48 hours just to create the radiation matrix.

luvMeSome142 & some Lincoln! > 64Mali
10/09/2015 at 10:17

Wow. Those are much bigger models that I normally work with. I do everything on a 3 year old 1.73GHz Intel i7 PC with 16GB ram. I’ve only had over a million nodes in a few models.

I’ve done one small project that involved cfd and radiation. It did take a long time to do the radiation calcs.

64Mali > luvMeSome142 & some Lincoln!
10/09/2015 at 10:20

I was doing military aerospace. Lots of boundary conditions, this was also 2007-2010 so computers were not as fast as today.

Kailand09 > luvMeSome142 & some Lincoln!
10/14/2015 at 17:59

Great suggestions. I’m pretty sure there are ASME meetings near, as my boss is a pretty involved member in both ASME as well as SAE I believe.

Never heard of meetup.com, though now I’ll look into it!