53 replies to this topic

[NeilR]

What sort of load ratings are acceptable for a car such as an F3/FF car at the upright and at suspension arm end? Actually does anyone know what size bearings and loads are acceptable to different manufacturers.
I thought the loads would be relatively modest and then I saw the replacement bearings for RALT/Van Dieman FF ( http://www.pegasusau...sp?Product=3075 ) rated at 28,500lbs (radial static) and with a half inch bore....much more than I expected.

[Greg Locock]

What sort of load ratings are acceptable for a car such as an F3/FF car at the upright and at suspension arm end? Actually does anyone know what size bearings and loads are acceptable to different manufacturers.
I thought the loads would be relatively modest and then I saw the replacement bearings for RALT/Van Dieman FF ( http://www.pegasusau...sp?Product=3075 ) rated at 28,500lbs (radial static) and with a half inch bore....much more than I expected.

Sizing the ball joints is usually limited by wear, and the strength of the housing rather than strength of the ball istelf. For instance the falcon uses 27mm or 28mm balls in the lower arms. I can't remember what the rear arms use.

The loads on the inboard ends of upper wishbones tend to be rather less than for the lower arm.

Here's the Aurora catalogue, it says your general loads should be 10-20% of the capacity of the joint. http://www.aurorabea...net_Catalog.pdf

Those bushes we found will have enough capacity for any of the locations on your car, as they take the full spring+shock+jounce bumper loads for a Falcon

[NeilR]

I wasn't thinking of changing Greg, I was more interested in what formula cars use and why.

[Joe Bosworth]

NeilR

Sometimes I think it is dangerous trying to do too much finite engineering.

In my experience the bearings outside of frame suspension points are the last places you want to save weight by going too minalmistic. How are you ever going to measure how hard you hit something before you accept a breakage? How often do you want to change things out on life cycle basis?

One of my favourite examples on this matter was a fairly heavy hit that I once had between my Brabham F3 car and a similar Lola. The Lola was constructed down and the Brabham was constructed for practicality. The Lola came away from the hit on the end of a hook after cleaning off both front and rear dangling things. The Brabham finished intact and post race check found all wheel angles unchanged. Both cars would end up weighing within 5 pounds of minimum weight and to one another.

Would you rather build to Tauranac standard or to Broadly standard?

Regards

[NeilR]

Actually I admire both of those designers a great deal, but they do come from different schools of thought. However let me clear this up: I am part of a group getting some small FSAE like 750cc hillclimb cars made up, though I will not do much constructing myself...and FEA will NOT be part of the process! The rating of the ball joints will assist as to whether we pay $18 per bearing from the US or $3 from China. The former is rated at 34kN and the latter 21kN and when you want 120 bearings it is worth investigating is it not? All bearings will be in double shear.

Edited by NeilR, 05 January 2010 - 07:24.

[cheapracer]

The rating of the ball joints will assist as to whether we pay $18 per bearing from the US or $3 from China.

My recent decision for this exact scenerio was easily resolved, I went for the next size up at $4 each to cover myself.

I probably added 1kg to the whole car but you should see the size of 'em, they ain't going anywhere fast.

[NeilR]

Thanks, we had the same reasoning...but 17-20mm spherical bearings is getting more than a bit silly! So finding out what is acceptable seems wise at this point.

[mariner]

The old advice on rod ends was to go down on size and up on quality. That way you still get the strength you want but with less weight, not just in the rod end but, potentially, on the associated tubes/cast brackets etc. The smaller size partially reduces the price premium of the better quality. Also which part will be the strongest on the suspension linkages? Does your load rating mean the joint will pull out or will the tubes fail first? I think without wanting sounded pedantic that the "best " scenario is to aim for deformation of the linkages whilst remaing attached so as to avoid wheel seperation if you can and have some chance of the tyres still retarding the car in a shunt. That might be the strees test case for the joint size but I am not sure.

A practical question is how you intend to protect the spherical load surface, it needs to be kept clear of all grit etc so you need to either keep it greased or fit covers. The load capacity won't last long , nor will the alignment if wear is rapid.

[NeilR]

True, but unit pressure is reduced with larger area and thus wear rate is reduced relative to smaller bearing. Bearings are PTFE lined and wear less. simple dust covers will suffice.

[Greg Locock]

The Aurora catalogues bears some scrutiny. In particular look at the load ratings for all their different series, notice how much they vary. I'm a big fan of Aurora, they do everything from cheap and cheerful to $$$$.

For a circuit car work out the load in each bush for a 3g vertical, 2g longitudinal, and 1g lateral load (seperately) at the contact patch.

Then take sqrt (sum of the squares). That tells you the radial load you are likely to see, and obviously 1g vertical tells you what you'll see all the time. Hopefully you'll decide on a balljoint with a rating something like 5 times the first and 10 times the second.

I think a larger cheap balljoint is probably a better investment than a neato little aerospace one, to be honest. Or at least, you can save more weight and money elsewhere more easily.

[NeilR]

Thanks Greg. Weight is not that much of an issue really (you've heard this from me before!) in that even if we are way over we'll still be looking at 350kg and the 750GSXR engine I have will be sprightly enough for us beginners in single seaters.

Edited by NeilR, 05 January 2010 - 11:32.

[NeilR]

Actually Axial load is perhaps more of an issue now I have looked at Aurora's catalog!

[Fat Boy]

I've gone through the 'saving money on cheap Chinese bearings' thing recently. You really need to nail down the QC on them. My bean-counter went cheap and ended up with 500 bearings whose OD moved around 0.002. The housing was machined at +/- 0.0005. It made for some angry customers and a bunch of cheap bearings that were unusable, which is expensive. I'm not saying all bearings from China are bad, but I am saying that you get what you pay for. Axial loadings are often a limiting factor, especially on wear. Make sure those numbers are conservative for a customer car.

[Bloggsworth]

The ball, whatever size, is likely to be strong enough, it's the threaded bits that bend or break. After picking up the pieces after many a crash, I can't ever recall seeing a broken ball or housing. Plenty of bent threads and bolts though. IIRC the Lotus 25 used 3/8" bolts all over the place, and the bolt through the inboard ball lower rear link of a Lotus 30 was 1/2" diameter, the only broken rear link I saw was on the ball joint, where the threaded part snapped off. John Dean came into the pits saying "It seems to be like jelly at the back end" - Pretty accurate as the lateral loads were being absorbed by the two doughnuts in the drive shaft.

Edited by Bloggsworth, 05 January 2010 - 19:14.

[Greg Locock]

So Neil, are you using cross axis ball joints, rod ends , or tierod joints with tapers?

[Lee Nicolle]

Very rough and ready. On a small car like you are playing with a 1/2' should be plenty. Or3/8 in aircraft quality ones. Though a good teflon lined 1/2"is cheaper and the pickup and bolt will be far stronger too.Lined is essential ofcourse as the metal metal ones rattle about the second time you use them!
Though I have bought rodends that have looked and felt good but lasted 1 meeting and have bought rodends for the same money that have lasted 15 years on a dirt car!
And last make certain you are not binding them, that is when they break.Check very carefully your full travel with no springs attached and then give it a bit more. 1/2 " is better there too as they have more misalignment angle, and you can buy high misalignment ones too. If using as a steering tie rod end be very careful as a normal tie rod end will cope with far more travel.Be prepared for the occasional off where everything flexes. That is what breaks the cages.

[Joe Bosworth]

Tauranac used rod ends with 1/2 inch eyes and 1/2 inch threads into 3/4 inch OD tubing members for the suspension bits.

I can´t think of any good or compelling reasons to go much smaller though if you calculate long enough you might convince yourself that even 1/4 inch rod ends will do.

Regards

[Greg Locock]

On a 250 kg car that was driven reasonably gently on public roads (but had to cope with rail crossings and potholes and so on) we got away with 3/16" or 5mm threads, but we did use the high spec Aurora ones not the standard range That also survived the belgian pave at the proving ground. That wa sthe last solar car I was involved with that went througha full development program, so we could afford to take things a bit closer to the edge. Therefore I'd be comfortable with a 10mm thread on a 500 kg car unless it was an off roader, and 12mm seems like a good choice for the upper and lower balljoints on the spindle on a double wishbone.

[NeilR]

Thanks to all for your replies. Current F3 cars use very small rod ends...perhaps not the best lead to follow.
Fat Boy thanks for the heads up, I will take that QC issue into account. We are aware of articulation angles and will be keeping some angle in reserve...though suspension movement will be modest. I have sourced two ex-CART/INDY steering racks from fleabay and we will need to find more or make some.
Greg, we hope to be using spherical bearings pressed into a machined housing on the suspension arm and no rod ends as such. Camber will be adjusted via shims at the upright end. We have discovered a rear bearing from a production car that has a live spindle and reasonably good bearing spacing.


We'll be using this at the front in a simple fabricated steel upright along the lines of:

Edited by NeilR, 06 January 2010 - 00:56.

[cheapracer]

I've gone through the 'saving money on cheap Chinese bearings' thing recently. I'm not saying all bearings from China are bad,

Hard to seperate the 2.

Next time give me a yell and I'll point you, theres a broad range of choice here.

The old advice on rod ends was to go down on size and up on quality.

Your statement is application dependent, I need minimul compliance but long life while bashing through potholes - cheap and large is my answer.

Edited by cheapracer, 06 January 2010 - 11:15.

[NeilR]

Do you know of a good company Cheapy who make very high quality stuff. All I've found so far are: http://www.unifor.cn/cpjs3.asp

[cheapracer]

Do you know of a good company Cheapy who make very high quality stuff. All I've found so far are: http://www.unifor.cn/cpjs3.asp

I have contacts for these things yes - sometimes there is a quantity issue though.

Tell me specifically what you want and I'll ask.

[NeilR]

Here you go. I would like to know if I can get and the cost of each of the following: GEG8T/X, GEG12T/X from:http://www.unifor.cn/cpjs3.asp
Do you know the companies in Cn that make the aerospace stuff?
I'll contact the company and see if I can get a sample too.

Edited by NeilR, 08 January 2010 - 01:31.

[cheapracer]

Here you go. I would like to know if I can get and the cost of each of the following: GEG8T/X, GEG12T/X from:http://www.unifor.cn/cpjs3.asp
Do you know the companies in Cn that make the aerospace stuff?
I'll contact the company and see if I can get a sample too.

My occasional drinking mates are http://www.sigmacomponents.com/ who make a few parts, mostly fuelline stuff for Rolls Royce and being Friday today ......

I'll ask about the above mentioned company today.

I will see my Mate who makes the Cold Isostatic Presses on Sunday and he was educated in Germany and knows what I need when I say quality and has a line on just about anything.

For a different purpose I bought some good quality brass ones just yesterday not that they will suit you but they were only $2 each! - not ideal for long term suspension use probably.

[cheapracer]

By the way, whats the hub from, looks smaller than a Honda one....

[NeilR]

later model Toyota Corolla.
As you note life is an issue and there would have to be at least two or three good manufacturers of such bearings in China

[Catalina Park]

Hey Cheapy, while you are looking up bearings can you try to find a substitute for the Timken IR-27 (used to be Timken or Torrington 3W5-8)
It the the clutch throw out bearing for a Mini. (it is the proper one and not the useless one that most places sell these days. )

[NeilR]

Isa this one of the new mini's or the old BMC one?

[cheapracer]

Isa this one of the new mini's or the old BMC one?

Old no doubt, haven't you seen his twitches?

Cat Park, that number strikes no references in Google - I may need dimensions.

And thanks for reminding me of the one time i did a clutch on a Mini, getting a twitch myself now.....

[Catalina Park]

Old no doubt, haven't you seen his twitches?

Cat Park, that number strikes no references in Google - I may need dimensions.

And thanks for reminding me of the one time i did a clutch on a Mini, getting a twitch myself now.....

Twitch? They are not twitches. It is hyperactivity. As for changing a clutch on a (real) Mini what car could be easier? I can do a clutch in two hours including a leisurely lunch break. The only car that is easier is a Holden Camira.

The dimensions for the bearings are 15.8 x 48.8 x 15.8

Here is a couple of photos, I hope you can see them...

[gruntguru]

Hey Cheapy, while you are looking up bearings can you try to find a substitute for the Timken IR-27 (used to be Timken or Torrington 3W5-8)
It the the clutch throw out bearing for a Mini. (it is the proper one and not the useless one that most places sell these days. )

Hi C.P. This takes me back about 30 years to when they first started selling the crap replacement bearing. Fortunately the old style was still available, because the crap replacement had a very short life when using any sort of serious clutch clamping force - in fact with two racing diaphragms stacked, they wouldn't even survive one application of the clutch.

Have you pulled a bearing apart? I seem to recall the original bearing is an angular contact ball and the replacement is a deep groove ball type. The former type is capable of withstanding quite high axial loads and the latter much lower loads.

[cheapracer]

The only car that is easier is a Holden Camira.

For the non Ozzies who don't know what he is talking about a front wheel drive Camira's input shaft can be drawn from the gearbox externally enough to let the clutch and pressure plate drop out without removing the gearbox. Great idea.

Of course if only the other 99% of the Camira had been as well engineered .........

I actually got a write up in Modern Motor after I drove a brand new Camira from Brisbane to Melbourne before they were released and my thoughts. Driving a light car with too much compression damping down the potholed Newell Highway for 12 hours was what led to my commenting on the car.

[gordmac]

Vauxhall/Opel used that system some time ago, presumably the same engine/box.

[cheapracer]

Vauxhall/Opel used that system some time ago, presumably the same engine/box.

Yes as it was the "world car" (J series I think?).

What a shame the first 1600 was a pile of crap, the later 2 litre was a much better and embarrassingly fast but the early ones had tainted it.

[Lee Nicolle]

Vauxhall/Opel used that system some time ago, presumably the same engine/box.

The GM J car was sold as Vauxhall, Opel, Caddilac!! and Holden Camirs. Powered by the Holden Family 4 unit which has all sorts of capacitys used in J car, Aussis Nissan Pulsars, and Daewoos.
As Cheapy said the original engine was junk. The 2 litres go quite well though a little harsh and noisy. Have some great ideas that backfire, The water pump is concentric so can be used to adjust the cam belt. But ofcourse they are always corroded in so you lose the pump everytime. Some these days will not come out at all. Cheaper to replace the engine than smash the pump out etc. Most of the cars are worth next to nothing so a simle watwer pupm or timing belt replacement writes the car off!!

[cheapracer]

Here you go. I would like to know if I can get and the cost of each of the following: GEG8T/X, GEG12T/X from:http://www.unifor.cn/cpjs3.asp
Do you know the companies in Cn that make the aerospace stuff?
I'll contact the company and see if I can get a sample too.

Ok, the GEG8T and GEG12T are about 10 and 14 RMB each and the GEG8T/X and GEG12T/X are about 30 and 40 rmb each with min order of 10 for each model plus freight from another province to me plus freight to where ever you are.

rmb = Chinese currency, USD is about 7 rmb to $1.

[NeilR]

Thanks for this, were you able to find out if there was a better quality manufacturer?

Edited by NeilR, 11 January 2010 - 04:27.

[NeilR]

I've gone through the 'saving money on cheap Chinese bearings' thing recently. You really need to nail down the QC on them. My bean-counter went cheap and ended up with 500 bearings whose OD moved around 0.002. The housing was machined at +/- 0.0005. It made for some angry customers and a bunch of cheap bearings that were unusable, which is expensive. I'm not saying all bearings from China are bad, but I am saying that you get what you pay for. Axial loadings are often a limiting factor, especially on wear. Make sure those numbers are conservative for a customer car.

It would appear from the chinese catalog that the OD can vary up to .008mm...were you measuring "/mm?

[Greg Locock]

It would appear from the chinese catalog that the OD can vary up to .008mm...were you measuring "/mm?

How are you machining the housings to better than half a thou?

[NeilR]

How are you machining the housings to better than half a thou?

I was going to go to Ross the engineer "an interference fit on this please" and then pick up the bits! Since his business is certified to do aircraft work I'd hoped to leave it in his capable hands (he is likely build the same car and will be using the bits too)...so in short I have no idea if they can. I can ask if you are really interested, but I presume that they send precision turning out to other businesses in the area if they do not do it themselves.

[cheapracer]

How are you machining the housings to better than half a thou?

Precision grinding most likely - or every racers friend, Locktite!

Edited by cheapracer, 12 January 2010 - 03:17.

[gruntguru]

I was going to go to Ross the engineer "an interference fit on this please" and then pick up the bits! Since his business is certified to do aircraft work I'd hoped to leave it in his capable hands (he is likely build the same car and will be using the bits too)...so in short I have no idea if they can. I can ask if you are really interested, but I presume that they send precision turning out to other businesses in the area if they do not do it themselves.

I think Greg is suggesting that a tolerance of 0.008mm (+/- 0.004mm) is OK, when your interference fit is probably about 0.02mm on a 20mm diameter bearing.

[Fat Boy]

It would appear from the chinese catalog that the OD can vary up to .008mm...were you measuring "/mm?

I'm 'Merican, Neil, I was talking inches. I would have turned back-flips if they were +/- 0.002mm. It was amazing really at how bad the bearing we got were. The OD was ground. How does a computer grind something with that big of a tolerance? I honestly didn't know that it was possible to F a bearing size up that badly. Lord only knows what the balls looked like. The probably used jelly beans.

[NeilR]

Gruntguru I suspected as much but I though I'd play it with a straight bat...cricket season is here! Summer in Australia, working in the garage, cricket on the radio (best place for it) and a cold beer...heaven ona stick!
Fat Boy I thought so. When do you think the US will go to metric? It would be a lot easier you know.

[Tony Matthews]

When do you think the US will go to metric? It would be a lot easier you know.

I think they are waiting for everyone to go Imperial. Some years ago I did a small job for Lola Cars, a rendering of a new F3000 car. Not a lot to go on bar an unfinished, very small wind tunnel model, and a profile of the latest version of the nose. This had staions marked on it, and a datum line. There was no indication of scale, and without bothering to scrutinise it I said "Er, is this in inches or millimetres?" Eric Broadly, for it was he, backed off as if I'd threatened him and said "MILLIMETRES?" Oh, that'll be inches, then.

[NeilR]

I was told recently that there are three countries that have a system other than metric: USA, Myanmar and Liberia...I have no idea if this is true.

[cheapracer]

I'm 'Merican, Neil, I was talking inches. I would have turned back-flips if they were +/- 0.002mm. It was amazing really at how bad the bearing we got were. The OD was ground. How does a computer grind something with that big of a tolerance? I honestly didn't know that it was possible to F a bearing size up that badly. Lord only knows what the balls looked like. The probably used jelly beans.

Ex State machinery and/or pathectically trained labour.

[Fat Boy]

Fat Boy I thought so. When do you think the US will go to metric? It would be a lot easier you know.

We'll get there....inch-by-inch.

In all fairness, since it was a metric bearing on an otherwise imperially measured drawing, I called it out in dual units. I easily operate mathematically with metric units, but can't estimate them worth a damn. Weird.

[Fat Boy]

Ex State machinery and/or pathectically trained labour.

or both?

[Tony Matthews]

or both?

...and/or...