Differential Material? 5

[Timms]

I am trying to get an idea of what material to use if I want to build a differential for a very small vehicle? Does anyone have recommendations? What do most automobiles have?

Blessings

 

[Pressed]

In the long term, this might help.

http://aplsweb.com/Topics/Differential.htm

_______________________________________
Feeling frisky.........

http://www.tailofthedragon.com

 

[Timelord]

How small is your vehicle? I bought a small differential for a go-kart that I built for my kids.
See it at:

http://www.hoffcocomet.com/comet/oem-differentials.asp

Timelord

 

[NickE]

4340-or-4140 (4130)

Seriously a good alloy steel will work great for the gears, and teh case could be aluminium or plastic. The trouble with building your own is that it is almost never cheaper.

 

[Timms]

Thanks alot for your responses! I'm not too worried about the cost. I'm actually designing rear-ends for motorcycles. We are converting bikes to trikes and I need a good idea of the correct material to use for a pumkin (differenctial housing).

Blessings

 

[Frederick]

8620 makes a good gear material, 9310 is better. You can get into exotic stuff but I don't believe your application warrants it unless you are trying to reduce weight or size.

For maximum strength and durability, vacuum carburize the above materials and give them a real -320F cryogenic process with proper ramp down, hold and ramp up, followed by a proper tempering.

 

[Timms]

Frederick

Thanks for giving me details. Do you have any suggestions for the carrier material. I know ductile iron could be used but I also heard that aluminum alloys work well too. I was thinking of a AA 356 sandcast or a 6061 T6?

Blessings
Drew

 

[thundair]

I agree with Fredrick and have used 8620 in carburized condition for a gear box. Also used it in a PSRU for an aircraft Mzda engine. The carrier could be made with lost wax casting if you are using 3D model definition that is able to transfer to a STP or STL file and the caster can use SLA as a pattern and 5 to 15 days later there is your carrier.

Regards

 

[carnage1]

I would caution you aganst using a cast aluminum housing, but a housing made from extruded 6061 and welded should work well and reduce weight.

 

[Timms]

carnage1

Thanks for the warning. What are your concerns with using a cast housing?

Blessings

 

[NickE]

I would like to know why carnage1 doesnt reccommend a cast housing also. Considering that most if not all production differential housings are cast.

 

[Timms]

NickE

I should have specified aluminum casting in my reply to carnage1. I think it is the "aluminum" casting that he is cautioning me against.

Blessings

 

[GregLocock]

I can think of at least one cast aluminium diff that is good for 300 hp.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[carnage1]

A cast aluminum diff could work but I would be very cautious. I have not cast any aluminum parts. my experiance comes from machining. none of the cast aluminums I have worked with were nearly as good as 6061 extruded. the cast aluminums tend to have almost no spring to them so they either permanantly deform or start stress cracks, depending on how soft or brittle they are. Extruded aluminum seems to be able to handle some overloading without failure a lot better than cast aluminum.
As for cast iron it is great stuff,a bit heavy, but it never warps. (unless my grandmother has been cooking with it)

 

[Timms]

carnage1

My only problem with extrusion is the price of material will be more and the process of machining will be quite extensive. A cast part comes already hollowed out and close to the dimensions of the finished product leaving only a little left to machine. The die will cost a fair bit but we plan an making alot of parts.

Blessings

 

[NickE]

If you plan on making "alot of parts" then definately a cast housing will be more cost effective. Also a proper design of the casting can produce a part with optimized geometry and gussets so that it is effectively stronger than a machined and welded part.

 

[carnage1]

maybe an extruded tube for the center and cast front and rear covers? It sounds like cast may be the only cost effective option if you are doing a large number of units.
Could you stamp a thick enough aluminum to make the rear cover out of?

 

[NickE]

carnage1- I know that you have some experience with this, but I beg to differ. If he is going to make "alot" of these, the best method is casting.

Many people think that casting makes brittle, failure prone components. They are wrong, just tell GE or PCC that casting is not the best method to make turbine blades and they'll tell you to jump off a long dock.

The steering knuckles in most (if not all) of the production vehicles in the US are cast. They are not brittle; in fact quite the opposite.

An A356-T6 sand cast housing would be the optimum, there is geometry that can be cast in that will provide plenty of stiffining and strength. Once the patterns are made and tested the cost per pound drops spectacularly. The rear cover could easily be die cast/permanant mold cast. (at least thats how the one on my car is made) I dont know about front cover, mainly because I've never seen an automotive diff with one. (well on front axle, but then it doesn't have a "rear" cover)

Sure a highly engineered machining from billet 6061-T6 would be awesome - the costs would be through the roof. So would the swarf. (or the production efficency would be really low.)

Also welding on 6061-T6 is only possible if you are going to re-solutionize and age after welding. (otherwise the weld is going to be really soft.)

Teh cover could be fine-blanked, but if you think molds/patterns/dies for casting cost a pretty penny, you should see what a good stamping die costs!

 

[carnage1]

Alright I give, it sounds like the only plausable solution is casting due to cost.
aprox. how many units does it take to make casting break even?
or are there multiple types of aluminum casting with various setup costs?
can you weld annealed 6061 and then heat treat or does the entire piece need annealed again before heat treatment?

 

[GregLocock]

For things like diff housings I'd have said a wilely guy could make a cost effective sand casting for ones or tens off. I had a cast part made that cost, for a one off, less than the welded proto cost, including the pattern. The reason is that with a sandcasting you can make your pattern any old how, including basing it on an existing part. You'll have to include draft angles and allow for contraction, of course, but for a complex shape an existing part makes a great basis for the pattern. Talk to or work with the pattern maker at the foundry.

Commercial automotive components are more likely to be low pressure die cast. There is a slight improvement in porosity and shut lines, but if you can afford the time to remelt the failures, there is not a whole lot wrong with sandcast, as, in this case, stiffness is more important than strength. I'd say that in the case of wheels, where strength is more important, that I would be wary of sandcast parts.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.