Ball joint taper tolerances?

[pmbrunelle]

Hello! I've been a long-time lurker on here, but finally I have a question of my own.

I have to come up with a steel part (turned on a CNC lathe at a supplier) that resembles a male ball joint stud. It has a 14.25 deg included angle, and there is a nut on the end to draw the male taper into the socket. Unlike a ball joint (three forces, but no moments), in this use case three forces and three moments are possible. Not sure if that changes anything with regards to the taper.

The female side of the joint isn't under my direct control, but I can transmit my requirements to those folks. The material of the female side is aluminium; this cannot be changed.

Looking in Machinery's Handbook, 0.001" taper error per foot seems to be normal, for Morse-type tapers in machine tools. Is this level of accuracy normal in automotive suspension, or is less accuracy typical?

My understanding is that the female taper tolerance should deviate as 14.25 +0/-X, while the male taper should deviate as 14.25 -0/+X. This would guarantee that in the event of an angle mismatch, contact between the tapers would occur near the base, away from the nut. If the contact were to occur near the nut instead, I imagine that large forces would be generated near the nut, and the female part could get squished / damaged. Is my understanding correct on this?

What is the correct way of inspecting the taper angle? So far, our part supplier has been using a CMM, but the CMM variance may be too much (if I'm shooting for machine-tool accuracy), and I'm wondering if we need to buy gauges (representing ideal female sockets), and do something with machinist's blue or similar.

I have some prototype parts on hand right now. Using a Sharpie felt-tipped marker, I drew a lengthwise stripe on the male taper, and I inserted the male into female (aluminium prototype part). With axial pressure provided by my hands, I spun the parts a few times back and forth. With this method, I could see if the contact occurred near the top of the cone, everywhere, or at the base of the cone. However, I re-did some parts (cleaning off the ink with isopropyl alcohol), and the results didn't repeat.

How much does any of this matter? Do aluminium female sockets simply deform and adjust themselves to the steel stud?

********************************************************************************

Concise version of my questions:
1. What tolerance is applied to the taper angle of a ball joint stud?
2. What inspection method is suggested?

 

[GregLocock]

Aluminium parts certainly do deform. I could find out what tolerance is used, but I couldn't tell you. However the manufacturers of tie rod ends probably have design manuals available. In general hard taper pins in soft sockets can be analysed using Lame's formula.

This document says you want 1mm pull through in aluminium

http://www.powersandsonsllc.com/pdf/tutorial.pdf

. You can back track on that to see how much deformation they are expecting, and hence your tolerance.


Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[pmbrunelle]

I assumed that with the ubiquity of ball joints that the design cookbooks would be everywhere, but apparently not!

I wasn't really wanting to devote too much study to this tapered fit as it's a small part of an overall bigger system, but I may be stuck working from first principles.

Our supplier is suggesting the use of a master gauge and Prussian blue, with a certain percentage of contact between the gauge and the part to be measured, so no more quantitative angle dimension.

 

[GregLocock]

It seems like a reamer gives a good enough solution for the hole so I suspect you are on the right track.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[pmbrunelle]

Poking around, it seems like the angle tolerance for automotive is about 10x bigger than for machine tools. I guess that ball joint tapers are also relatively short compared to machine tool tapers, so the Lame's formula stretch between the small and big ends of the taper doesn't vary too much.

70% (this value seems to pop up) Prussian blue contact seems to be another way of controlling things. Our supplier seems to favour the ink method, so I'll be focusing on that.

One thing that needs to be determined is the application of the blue ink to the master gauge. As the ink thickness becomes very thick, ink will touch all sections of the measured part, effectively passing all parts (not necessarily the desired result). So the ink thickness needs to be controlled.

Master gauges seem to be available for a few hundred bucks apiece, so we will probably order some gauges (some for my employer, some for our supplier). Hopefully ordering the gauges for all from the same gaugemaker prevents disagreements and helps with comparing apples with apples.

The only real technical question remaining is to decide what angle the master gauge needs to have. It will need to have an angle slightly larger than 14.25 deg if I don't want it to pass parts with an angle less than 14.25 deg.

If the small end of the part touches the gauge, and the big end of the part just touches the surface of the ink, I would need to make the gauge with this angle:
14.25 + 2*arctangent(ink_thickness / taper_length)

I don't know what the ink thickness is for a given ink application procedure, but determining that is next on my to-do list.

 

[gruntguru]

Machine tool tapers are designed to provide a high friction fit that can be seperated and re-fitted indefinitely. The primary functions are precise location and to transmit torque along the taper axis ie pure shear across the mating surfaces. The fit needs to cover a large surface area of the taper with contact pressures remaining in the elastic range.

Automotive ball joint tapers carry loads on at least five axes. They are typically tightened sufficiently for surface iregularities to deform plastically, so a large contact area will be achieved even with lower tolerance components.

je suis charlie

 

[pmbrunelle]

On my personal cars, I've had no problems with replacing ball joints and tie-rod ends, but I have only owned things with cast iron knuckles.

With an aluminium knuckle, does the tapered stud tend to sink into the knuckle with each install/uninstall cycle?

In service, I'm not expecting this widget to be removed very often, but in prototyping/development, I expect a lot of part-swapping to occur. Should I be on the lookout for the hole sinking in?

 

[SwinnyGG]

With an aluminium knuckle, does the tapered stud tend to sink into the knuckle with each install/uninstall cycle?

Not really.. but in a real automotive application, removing and replacing ball joints is a rare operation; a knuckle might only have the ball joint interfaces broken, and new ball joints installed, once or twice over its entire life.

In service, I'm not expecting this widget to be removed very often, but in prototyping/development, I expect a lot of part-swapping to occur. Should I be on the lookout for the hole sinking in?

If you're going to be swapping ball joints a whole lot in prototype parts, yes- you need to watch for dimensional changes in the female hole as well as changes in surface finish.

 

[GregLocock]

You could put a steel sleeve into the spindle for a prototype or race car.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[pmbrunelle]

In this scenario, the tapered joint itself represents a bunch of the technical uncertainty, so I want the joint to be tested (test drivers beating the crap out of the vehicle) using the production-intent materials.

When the tapered joint is not (among others) the element under study, then your idea of the steel sleeve makes sense!

Looks like we're going to standardize on Permatex Prussian Blue, non-drying. It's readily available, so anyone who wants the check the taper should be able to do so easily.

 

[pmbrunelle]

I could find out what tolerance is used, but I couldn't tell you.

So I ended up being able to find someone in my company who has direct experience with these parts.

One thing that I find difficult is the transfer of information. There are about 10000 people working for my employer, but on a regular basis, there are perhaps only 5-10 engineers in my surroundings that I can ask for tips/advice/opinions. The knowledge of the other employees is hard to tap into... I am sure that there is some degree of redoing things for the first time, even though it's not the first time for the company.

What I did to find the ball joint / anti-roll bar guy was to go through sales. Salespeople, being the client-facing frontend, are usually able to sell any product, so they have relationships to the technical people behind each product.

How do you all share knowledge among colleagues?

When I was in the armed forces, it was the opposite. You could meet a stranger wearing the same uniform as you and you would both know more or less the same things.

 

[GregLocock]

We have a wiki for our section, and a formal report repository, but yes, corporate Alzheimer's is a real thing.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?