I’ve seen a variety of posts reference different materials being used to support cantilevered wheels, like those commonly found in a west coast drive. Can anyone with experience comment on the suitability and successes or failures of 6061 alum vs 7075 alum vs 4140 steel?
Don’t use 6061 Hex Axles in a Competition Drivetrain. One of the teams in EWCP (I’ll let them chime in later) had a catastrophic drive failure last year due to one of their drive shafts breaking in two. (Axle was a 1/2" hex.)
7075 is what many of the West Coast Teams run, 2024 will work as well though it’s not as strong as 7075 (2024 is the Standard grade for AM extended hex shafts). The Majority of these teams run either 7/16" or 1/2" hex.
I’m not familiar with 4140 Steel in Drive-shaft applications so someone else will have to chime in.
We have been using 1018 carbon steel 1/2" hex for our shafts for 3-4 years now. It’s half the price of the 4140 and is more than adequete for FRC use. We generally don’t want to worry about axles failing and take the weight penalty of the steel down low where it helps the CG.
In our pre-2009 drivetrains that had dead axles we just used grade 8 3/8" shoulder bolts because it was an easy way to assemble them.
There’s a lot more to the question you’ve asked besides just what material to use. Size, stress risers, fatigue, etc. are all factors you should consider when choosing a material.
That being said, you can often look at similar applications success/failure to determine what may be a good place to start. Dustin’s advice is a good start. To add to it, we will generally use 4140 steel for our shafts. We’ve never had a failure, it’s a relatively easy to work with material (not that the aluminum alloys aren’t either), and it’s quite cheap and easy to find.
Obviously the big advantage to the Aluminum shafts is the weight savings. However, to put it in perspective… A 3/8" hex shaft of 4" length weighs ~0.137lbs when made in steel, and ~0.048lbs when made of aluminum. A total weight savings of ~0.089lbs. Like I said it is measurable, however even replacing 6 of these steel shafts with aluminum gives us a total weight savings of 1/2 a pound.
We’ve had plenty of success with aluminum shafts as well as the steel ones. Weight has generally not been a huge issue for us, so we take the extra security and go with the steel. The design priorities for your team most certainly are different so you should be sure to make your own evaluation based on our advice.
-Brando
I would not be comfortable with 6061 in any application really. This past year we bent 1/2" round 2024 keyed axles. We used 2024 1/2" hex on the gripper which was fine. I plan to go to all 7075 in the future. It’s much stronger. Fry Steel has good pricing on 7075 1/2" hex, but it may be a tad oversize. For round axles that you want to fit through bearings easily without any work, precision ground 7075 would be the way to go, but it isn’t cheap.
Material alone isn’t enough to answer the question; the overall geometry is very important. You could use wooden dowels for axles and be fine, and use very high quality 4340 steel and shear them every match.
We use a lot of 7075-T6 Aluminum (and rarely any steel), but we generally always do the analysis to see if these shafts will fail (and go to a higher quality steel if necessary, or increase diameter, change how they’re loaded, etc…).
If you have an engineer on the team, or someone willing to give you some time, they should be able to do these calculations for/with you (it would be difficult to teach students to handle it for all cases, but for some specific repeatable ones they could). If that’s the case, I’d say go ahead and look at aluminum as an option. For someone with an engineering degree, this should be trivial.
If analysis isn’t an option for you, I wouldn’t trust anyone’s anecdotal evidence really. I would stick to steel, preferably something real tough and strong like the chromoly series.
EDIT: A useful trend to know, even if you can’t do the actual calculations, is that torsional strength of a shaft goes with diameter^3, shear/tensile strength of a shaft goes up with diameter^2 and bending strength of the shaft goes up with diameter^4 (someone please verify these relationships for me, been a while since materials). Using this you can quickly realize/show that increasing diameter is one of the best things you can do for shaft strength. Bending is also the most likely failure mode for FRC shafts as well.
Fixed. I mean no offense, but this is the halo effect at its worst. I’m about to finish my second engineering degree and wouldn’t call the calculation ‘trivial’. Sure, I might know where to start in order to learn how to do these calculations, but they’d take much more time & effort than implied.
This was one of my main thoughts. For the added strength and relatively low weight penalty, why not just use 4140?
Follow up question, again as it relates to West Coast Drives: is it typical to just use snap rings or e-clips on the end of the shafts to hold everything together? It seems that there could be relatively significant side loads at times of heavy turning or impact. Has anyone ever seen a snap ring/e-clip give way?
I have seen this happen when the snap ring groove was not well toleranced, but that is about it.
We’re working on a WCD drive right now for the off-season (future t-shirt cannon, probably) and we figured that tapped ends, bolts and fender washers would be a better choice than little snap rings.
However, the only time I’ve seen one fail is in a Toughbox Nano, and then only after quite a bit of abusive “engineering”, so I wouldn’t be that surprised if it works for a drivetrain.
I haven’t seen a snap ring give out directly, but there was a case last year where a team had a snap ring groove in a torque-loaded section of the drive shaft. That is a serious stress-riser and it caused the shaft to fail.
A few thoughts: While 7075 Al has a higher yield strength than 4140 steel, 66ksi vs 60ksi, 7075 has a lower ultimate strength than 4140, 78ksi vs 95ksi. That means that the steel shaft will start to permanently bend at a lower load, but the aluminum will actually break in half first in a bending-load situation.
Aluminum and steel do not have the same tensile-strength to shear-strength relation. A rule of thumb is that for steel, shear strength = 0.75tensile strength, but for aluminum shear strength = 0.65tensile strength. This means that 4140 has a slight strength advantage in shear strength, the strength required to resist torsion. 4140 has the option to be heat-treated by most small machine shops to be considerably stronger, easily twice the strength of 7075 T6 aluminum.
I would use 4140 steel for the reasons I discussed above, and because I’d have the option to weld features onto the shaft if I wanted to.
Why do you feel it’s a better choice?
A shaft that yielded is just as useless as a shaft that broke.
I guess it’s just the idea that more metal means more secure. A bolt seems less likely to be wrenched out than a snap ring.
Not necessarily true. A bent dead-axle could still function and drive. A bent live cantilevered axle might throw a chain, but will still rotate. One might also get penalized if a part of the robot broke off and was left on the field.
Some yielding, especially localized yielding, does not render a part as failed. It was a hard mindset to get into after I got out of college. I now always think through “what if this part yields a little bit right here?” and evaluate if it will still function afterwards.
On the subject of failure, a steel shaft will deflect much less at yield than a similarly sized aluminum shaft despite having a lower yield strength. And either material will recover its shape after bending. In my experience forming both materials, I found that alloy steel will recover it’s shape (spring back) more than aluminum will.
Sure you can argue it either way, I’m just offering my point of view and my experience.
This is what we do to trap our axles and bearings in place in our drivetrains. But then we do live non-cantelievered axles making this a more logical setup.
Advice for that method: check those bolts before and after every single match to make sure they’re tight. We had one that managed to come loose last year, and it took the span of several matches to fix the damage caused (this was on a normal, run-of-the-mill 6WD, not technically WCD).
On the other hand, I used snap rings for three years with no problems other than “I can’t get the @&*#$^ snap ring off”.
Never have had any issues with snap rings coming off in a properly machined groove.
As someone stated earlier- the grooves are stress risers though so it could be the potential failure point on the shaft if it were to fail at some point.
Tapping the ends of the shafts is also something we’ve done before with good success.
-Brando
A 1/2" external retaining ring can take a 4000 lbf thrust load. No problems whatsoever as long as you cut your groove properly.
Agreed 100%!
Snap rings, with the correct groove, should do a much better job than bolts because of the actual loosening factor that occurs every match.