Wishbone Manufacturing

[Steve-Imperial]

Im about to start manufacturing the wishbones for IC02 and have a query:

Is it possible to purchase the threaded end pieces (also with spanner flats) welded into the ends of the wishbone tubes which you screw the threaded rod end adjusters into? If you can it would save us a load of time, especially since our undergraduate workshop is closed for another two weeks due to a workshop course!

Does anyone have any hot tips on wishbone manufacture that they might be willing to share? Our designs are fairly conventional with built in outboard spherical bearings and inboard rod ends. I envisage machining the spherical bearing housing post welding for a good bearing fit and a snap ring groove.

Cheers, Steve

[Steve-Imperial]

Im about to start manufacturing the wishbones for IC02 and have a query:

Is it possible to purchase the threaded end pieces (also with spanner flats) welded into the ends of the wishbone tubes which you screw the threaded rod end adjusters into? If you can it would save us a load of time, especially since our undergraduate workshop is closed for another two weeks due to a workshop course!

Does anyone have any hot tips on wishbone manufacture that they might be willing to share? Our designs are fairly conventional with built in outboard spherical bearings and inboard rod ends. I envisage machining the spherical bearing housing post welding for a good bearing fit and a snap ring groove.

Cheers, Steve

[fsae_alum]

Sounds like you're on the right track with the machining after welding and the snap ring groove. As for the threaded pieces you're inquiring about, try http://www.chassisshop.com/ They have them here in the states. Not too cheap though but they do work really well. They have right and left hand threaded versions.

[-]

I would agree that your on the right track for the bearings and the machining after you weld. It might be hard to fixture the peice with things welded on it, but I don't know what your peices look like so you might have no trouble at all.

As for the threaded pieces, we have a small lathe in our shop and I have found that it is much more likely that you will get a hole that is centered if you make a small bung that fits into the tube and then in the lathe drill out and tread the hole. This is also cheaper then buying the inserts from Chassisshop and you have a kid that is good on a lathe, he can get them done in no time.

Good luck.

[Denny Trimble]

Yup, inserts are a good way to break in the new recruits.

[jack]

i have been wondering how auburn made the gussets on their wishbones. im sure other schools have similair gussets. i could figure out how they are manufactured, or were they purchased?

[UTA racer rikki]

Jack,

Chances are, at least in our case, the gussets are made in house. A little extra material (and a good welder) and you can avoid post weld machining the bearing cups. We have tried it both ways here, and the results are similar. I would say stick to your plan of post weld machining though because it always seems to ruffle feathers when you change the process in process. http://fsae.com/groupee_common/emoticons/icon_wink.gif

As a side note, if you didn't hear what the design judges said at design finals, "We are going to try to keep any teams that have rod-ends in bending out of design semi-finals." If your goal is to do well in design, take that statement to heart and move away from using rod-ends inboard.

[Denny Trimble]

I thought I heard Jay O'Connell say that (what you had in quotes), but I also overheard him say that rodends on the inboard points are OK if they're aligned with the tube. Welded a-arms aren't true 2-force-members, but his opinion was that it's good enough.

I agree that inboard rodends aren't ideal on a welded a-arm, because if you're adjusting them at all, you're flexing the a-arm. Sphericals can be lighter, but your tolerances have to be better.

[MikeWaggoner at UW]

Aren't the inboard rod ends placed in decent bending loads under braking?

-Mike Waggoner

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Denny Trimble:
I thought I heard Jay O'Connell say that (what you had in quotes), but I also overheard him say that rodends on the inboard points are OK if they're aligned with the tube. Welded a-arms aren't true 2-force-members, but his opinion was that it's good enough.

I agree that inboard rodends aren't ideal on a welded a-arm, because if you're adjusting them at all, you're flexing the a-arm. Sphericals can be lighter, but your tolerances have to be better. <HR></BLOCKQUOTE>

[Storbeck]

I think the idea is that the loads can pretty much only go along the axis of the tube, so the inboard rodends would be nearly under pure tension and compression.

How do people without at least one rodend on the inside adjust caster? Or do you just not change it?

Has anybody tested to see what kind of bending load it would take to make a rodend fail?

[Daves]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>How do people without at least one rodend on the inside adjust caster? Or do you just not change it? <HR></BLOCKQUOTE>

An excellent example is UTA. They used shims, small slivers of aluminum to adjust the distances. They also used shims for chain tensioning, which is slightly different than most other teams.

[Colin]

we've used shims to adjust camber and caster for the past three years. it's good when track testing as we know how many shims equals what degree of camber/caster. Alot of ausie teams go this way from memory.

also we've found that if we machine the spherical housing first then weld it onto the wishbone and then clean it up with an adjustable reamer the spherical goes in nicely. because our the spherical housing is usally about 2.4mm thick and the wishbone tubes are only .9mm the spherical housing dosn't experience much deformation during welding

[IsheeM]

When you adjust caster by shimming the inboard pick-up points, do you change any of the other geometric settings?

[Colin]

no you don't and that's another good reason to use this setup. we've found that the judges seem to like it to

[Steve-Imperial]

Thanks for the useful advice people.

About the design of wishbones, I cant see why inboard rod ends (when aligned with the tube) is a bad thing at all! They represent a quick and easy way of making adjustments. As reccomended by someone on here im planning to make a spreadsheet which calculates the exact number of turns of adjustment to achieve different geometries. Faster than using shims I think.

[wacko]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Colin:
we've used shims to adjust camber and caster for the past three years. it's good when track testing as we know how many shims equals what degree of camber/caster. Alot of ausie teams go this way from memory.

also we've found that if we machine the spherical housing first then weld it onto the wishbone and then clean it up with an adjustable reamer the spherical goes in nicely. because our the spherical housing is usally about 2.4mm thick and the wishbone tubes are only .9mm the spherical housing dosn't experience much deformation during welding <HR></BLOCKQUOTE>

hi, havent not used spherical bearings for the A-arms. was just wondering wat did u guys used to keep the bearing in place? is it just a interference fit with a circlip holding it in?

thanks

[Colin]

we've used an interference fit one year and an interference fit with small punch marks to raise some material over the edges of the bearing, both methods have worked well for us, that's not to say that circlips arn't a neat way of doing it.

[Sam Zimmerman]

If you look at how Aurora does their rod ends, they use an interferance fit and punch the model number on both sides as a stake, just like Colin was saying. We will have a similar setup next year.

[BStoney]

There is a specific method for staking the Aurora Spherical Bearings into your "bearing plate" if you check out their website. It shows the correct depth for the chamfer on the hole based on what bearing you are using in addition to the dimensions for the tool to stake with. Check that out for proper manufacturing methods. Be careful with an interference fit on a bearing...don't want the case to bind with the bearing itself....

Good luck
B.J.

[Frank]

leave a bit extra, then weld

then machine on size after welding

just an interferance fit, and a dab of locktight for good luck

in our case the geometry dictates that they dont come out, direction of major forces etc

i use 002 interferance on teflon coated sphericals, and give them a "break in" spin in the lathe / drill to loosen them up after fittment

[Horace]

don't mean to bring up an old topic but can I ask how do you machine the press fit after welding the arms onto the housing?

Don't you need to lathe it? or do you mill it instead?

[Matt Gignac]

Either an undersized reamer, or a boring head. If there's a shoulder to hold your bearing in, a boring head is likely your only option. We made the holes 1/16" undersize, welded, drilled 1/32" undersize, then reemed to size -.001 for a staked joint.

Matt Gignac
McGill Racing Team

[Wizard]

Horace,

Check out this post on bearing installation from a while back.

http://fsae.com/eve/forums?q=Y&a=tpc&f=125607348&m=16410207621

[Christopher Catto]

Guys,

I used to think a lot about this welding and then post-machining technique but now that I work for a sportscar manufacturer I have realised it is a bit of an overkill.

Since your tube and your bearing housing are going to be relatively thin, all you need is to press a spacer in the housing while you weld to avoid the housing going a bit oval due to local heating. Then let it cool slowly (so that after heat expansion the parts return back to their normal size ) and press out the spacer. Steel does have some spring back but much less than say aluminium so this is why jigs or spacers are important to keep say 99% of the original shape.

Then any deformation which may have occurred will mean that the bearing will be slightly harder to press in, BUT due to the relative stiffness of the bearing compared to your thin-wall housing, the housing will deform and your bearing will not show any significant increase in friction. The bearing friction is generally related to manufacturing tolerances, manufacturer, age or hours of use, lubricant.

Our company uses snap rings but this is because our car weighs 1100kg and we worry about safety for the customers. For a race car, a press fit and a notch is sufficient. You can use also Loctite but have to ask them specifically for the correct gap in order for the glue to bond rather than be completely squashed out.

Really, spare yourself the machining and concentrate on getting the manufacturing technique fast and repeatable so that in future years you can pass the knowledge down. Invest your time to arrange with members of staff to have a folder where you keep all drawings and manufacturing processes for your staff to use in future. Documentation and repeatability is a key to success. How many times has a team done the same thing that was done 3 years ago simply because they did not know about it and no docs were available?

It is very wasteful of resources for each team every year to look at improvements which are actually just providing a 0,1% performance increase for a 10% extra time effort.

Go fix the most important things: drive drive drive. And log log log. Then tune tune tune.

[Drew Price]

Christopher,

I made turned steel press-in plugs when I welded our a-arm outboard spherical bearing mounts for last years car as an experiment. 0.625" x 0.049" tubing for the arms, 0.100" thick triangular gusset plate, 1" steel round stock bearing carrier, 0.75" dia spherical bearing, so a 0.125" wall thickness around the bearing. Not thick, but definitely not thin.

http://photos-a.ak.facebook.com/photos-ak-sf2p/v79/179/125/2413466/n2413466_32097208_5549.jpg


With no insert, I saw between 0.003" and 0.004" of housing out of roundness, and with the steel insert (which was about 0.0015" press fit for a 0.75" bore) I saw between 0.001" and 0.0015" of housing distortion after I removed the insert. The recommended press for these bearings into a steel housing is 0.0008"-0.0014" (from Aurora). Since the distortion with the plug was greater than the tolerance of my press fit, and since I pressed one in for testing anyway (it only wanted to rotate in one axis, where the bore was widest, and bound in the other) I decided to just make them 0.040" undersize, and finish machine with a flat bottomed boring attachment on the mill, left a positive stop shoulder on the bottom. I machined the snap ring grooves while on the lathe 0.020" deeper than I needed them so that when I had finished machined, they were the correct depth, and I didn't have to try to deal with doing them on the lathe.

So my question to you is, what is the relationship to the housing thicknesses you are talking about versus the bearing bore? Are you talking about steel, or Al, because 0.0015" out of round on an Al housing will have drastically less spring back than on a steel one.

Best,
Drew


EDITED: Photo no work.

[Drew Price]

Bump.

No one? I would really like to do this differently next year, or see what I did last year that gave me so much deflection.

I have already been told to use brass, or Al for the insert, and to make it bigger and or heat-sinked to pull head away from the bearing housing to keep the temps down in that area, but has anyone measured what kind of an effect that has? Any other ideas?

Best,
Drew

[Christopher Catto]

er. sorry for not replying to your question.

ok, so my experience is the following

1)
UH car 2004 had bearing cups with about 1,5mm wall thickness. i know, very thin. guy who did it was a good welder and he did tension compresson tests. that car also had wishbones with half inch dia tubing with again something ridiculous like 1mm wall. judges did not like it but i can say that the wishbones would survive racing. that car was raced the year after in class 1(200) in the UK. it failed the enduro because of engine cooling.

dont know about the deflection of the bearing cups after welding but i know the bearings were moving easily. however, i do not know details of the press fit

2)
for 2003 class 1(200) car i made some machined housings which then were welded into the tube. so, a bit like a female rod end, just without the thread inside. design was a bit complex because 2 of these housings were then welded together for the outboard bearings. was very heavy and un-necessary. kicked myself in the foot for all the over-engineering.

3) for the UH 2004-05 car, my classmate Jim made some laser cut 4 mm plates with a cup that would drop in (laser always leaves a notch at the start/finish so file this off. clearance should be 0,2mm) and be TIG welded around the circumference. it was perfect. really nice and used in F-3 and other formula cars. this will avoid asymmetrical distortion.

4)
where I work now we MIG-weld (yes!) the bearing cups to the tubing. tubing is 36x18 aero profile with 2mm wall (our race car has 500HP so needs some decent wishbones). the cups have 36,5mm outer dia and the bearings have 29mm outer dia. so wall is approx 3,75mm.

There are many different takes on wishbones. If you are really clever, you can simply get your welder to weld a bead around the circumference of the cup or just melt the edge (with TIG this makes a bit of a fillet). ok, so I may be going crazy, but simple things like this affect heat distortion.

when I now sometimes make chassis jigs (for various mounts and plates to be welded to the chassis) in order to stop things bending like a banana, i simply make a large weld pool on the opposite side of where I weld. This makes the weld area roughly symmetrical. As long as you have good fit and bolts clamping the jig, this is all it really needs to stop it springing back much.

Hope this helps. Just take a balanced approach so that you can make your wishbones easily and cheaply, at the competition if necessary or just make 3 sets (2 as spare).

try a few machined cups of different fit. weld them and then try a bearing. practice. bearing costs about 10 bucks so buy a few extra to use for R&D. it is always easier to remove pressed in bearings than ones held by Loctite. Loctite is nice for some things but I have learned to avoid it for sphericals. Too much worrying about the gap, the cleanliness, what if you need to remove the bearing (will it get damaged)

sorry for long post http://fsae.com/groupee_common/emoticons/icon_frown.gif

[Drew Price]

It seems to me that what might be happening is that the thinner wall thickness around the sphericals is plenty strong to retain the bearing, and even if it distorts some with welding, there is just not enough material around the bearing to squeeze it hard enough to bind? As opposed to the thicker housing, which has plenty of meterial to crush the bearing pretty mightily. Does that seem right?

Will have to analyze that a little more this year. Thanks!

Best,
Drew

[Christopher Catto]

yeah. you got the picture. it takes a bit of trial and error.

main thing is to have a design where you weld around a COMPLETE circumference. this avoids most distortion.

so a cup inside a cup is a possibility. i have seen this on race cars. anything that avoids machining is usually quicker.

i do not know what judges think about this. they may well prefer your idea of post-machining. so maybe ask some of them?? its one of those things where we have an opinion, they have another.

Anyway, good luck with the project. Remember not to spend too much time on it. Move on to other interesting things too!