Model Railroad Wheels

Sounds like a good plan.

Just a word of caution about differential thermal expansion of Al and unfilled epoxy resin.

The mismatch can be significant over 4.5 inches. I'm not worried about hot weather so much as cold weather. Resin shrinkage over your temperature range may be a problem. You may want to consider a filled resin for better thermal properties.

You would do well to study the relationship between cure temperatures and service temperatures.

In other words, if you cure it at 150°F the stress is more or less neutral at that temperature. If you threw it into freezing water the epoxy may just shrink away from the Al. Cure it in the middle of its service temperature profile.

Just a thought. Good luck.

I contacted Smooth-On, a company that manufactures rubber and plastic resins for casting. They recommended a mineral filled urethane resin of durometer, Shore 82-85D. Pot life for the 85D resin is 20 minutes and can be demolded in 2 hours. Shrinkage is .0006"/", tensile strength is 3100psi. Sounds like the right stuff. They claim it is tough and fully machinable. Based on what I have learned on this forum and Smooth-On, I will proceed on a trial basis and evaluate the results.

Good show. Send pictures.

GA

Although I don't know my way round such resin casting, all you said made very good sense to me personally. Also you put it well in a simple, fully understandable manner, thanks.

Well, depending on the load, I'd be a little concerned about such a thin 'tyre' (tread) deforming and growing, so de-laminating.

Secondly, such a thin tyre would require a very accurate location and circularity of the parted off tube pieces in the casting jig.

You have both prep machining and finish machining - which is fine if it's a labor of love.

But, if you are thinking of making and selling, I'd suggest you explore making a pattern and casting the 'minimum material' part in iron.

I agree with 34point5, I would be concerned about the thin 'tyres'

You could overcome this by starting with a much thicker walled tube, and I would also be concerned about the durability of aluminium for the tyre regardless of thickness.

I would suggest using thick walled steel tubing.

How were you intending to attach the hub to the axle?

Novel approach, but I suspect that 3D printing / manufacturing will eventually be even cheaper. It's being done in plastics and other materials now (which may then be used for metal die casting) but some companies are starting to work directly in metal.

http://www.gizmag.com/go/2578/

A 3/4" x 5" x 12' aluminum flat bar would cost about US$150 and could be made into about 28 solid wheels with simple machining (cut into squares, turn them round, drill center, and machine the indentation and flange). Just an idea.

True. I'm trying to avoid excessive material removal. My wheel casting is an experiment. Other wheels I plan to make are around 8" diameter and spoked. This will require a female mold made from a male pattern which will in turn be cast with resin, allowed to cure and finish machined on a lathe. Iron castings for spoked wheels are very expensive; much more so than solid cast discs. BTW, the rim and center core of the wheel can be steel. It doesn't need to be aluminum. I just chose aluminum for it's machining ease. I've designed various jigs and fixtures to accomodate the manufacture of wheels at each step of the process.

You can buy 5" or 6" 6061 AL or steel, slice them, and have very light machine work. All much cheaper,, faster, better than composite construction, I think. But I'm known to be wrong from time to time. (Just ask these guys) Good luck.

Slicing 3/4" pieces of 6" round bar is time consuming. I can via a jig, slice off 3/4" wide bands from 6" OD tube in a lathe, face it; pour the resin; finish machine after it is fully cured. I've ordered some resin and will try it out soon.

GA

Many of these supplies have a cutting service; water, plasma, oxy, and commonly supply discs out of plate.

Call any machine shop and ask them. If you want round plate, it's more efficient and cheaper to slice a rod, if the size is even remotely close. In this case, it could be exact.

How do you propose they "slice" a 5" or 8" rod and have the result as parallel and square as a disc out of plate?

If slicing rod it was 'more efficient' - it would be how they make washers.

"If slicing rod it was 'more efficient' - it would be how they make washers."

If one needs to make 3/4" thick washers, then yes, it is more efficient.

Cutting 5"+ round bar into discs is much easier & faster than machining rectangular plate. I agree that it will not be "as parallel and square as a disc out of plate", but a properly setup bandsaw or cold-saw can maintain tolerances within a few thousandths. If the part is to be roughed on a lathe, then a cut-off tool might be the best answer.

Drop a rod in a band saw, slice the discs. a typical machine shop might call for 20 thousandths over in thickness. If the diameter is right, the face work makes edges perpendicular. It needs light machining, could be all lathe work as I read it. Chuck the blank, face it, edge spec it, flip it, do the same. Maybe it's quicker to do the center hole work on a mill, but I doubt it.

If you saw from plate, you (would) will have to lathe the edges. but there is no production value. Aluminum is like butter to a machine shop. . Both methods require the same machines. I just went through the same excersice getting production 8" aluminum disks 1" thick with quite a bit more machining, but still, every shop I spoke with quoted 8" rod as the base material.

I see my approach is drawing flack.

Ok, say we take it as an aluminium product.

Typical billet will have the usual striations, and the first chuck setup can be a fiddle if the saw cut is not dead square. It may be a light first facing cut or a significantly intermittent one - so you have to allow for min two passes.

For the second cut, the blank has to be either on a face plate, or snugged to the chuck face (or jaw steps, if the chuck is accurate enough)

Now you can face, profile and bore the center (3 operations in 1 setup)

So in 3 setups (saw, face, face/drill) and 4~5 removal processes, you now have the same as a water cut 'washer' out of flat bar, or plate. (1 setup, 2 removal operations)

You could put the water-cut washer (variously) in a lathe and finish the bore, hub and facing profile (per OP dwg), then all to be done is the tyre/flange - via a hub bore jig and live center clamp - or drilled holes, face plate and tail stock cone centering.

Or;

You could setup two stages in your CNC mill.

1. in vice - drill, bore, profile hub/face (do spokes, if required).

2. on 4th axis, or hub clamp jig - tyre/flange profile inc flange back.

If you want, you can invert and relive out the back (i.e. a 3rd station if you have the table space for it) - (not shown in OP sketch, but 'anticipated' in typical hub/axle arrangements)

Guessing on one of my typical CNC mills - programmed as a 'manual re-jig';

Door closed to door closed, I'd estimate 45, but say 60 seconds. Or one finished product per the OP sketch profile, per minute. 60 per hour

But equally, given one good face on a sawn disc, you could do the same.

But: you would need the extra station to clean up the full back, not just the back of flange edge - and you'd have to add the time to drill, prior to hub boring - that a water-cut washer obviates.

I could do you a little Excel by way of comparison, but essentially, it's obvious where the cost benefit analysis 'breaks even', given a multi-head water-cutter dies n at the same time.

As is costing wise; if the labor component is equal, or greater, than the base material cost - then they/you, are doing it wrong - or you are locked into 'technically antiquated' suppliers.

Mind you, the above is about your case, that infers an 'authoritative stance' in a quantity situation, so charging for labor, overhead recovery and seeking the maximum profit the product will competitively support - i.e. not 'a hobby'.

I can see that you have never cut a circular blank with a band saw, mill or lathe. Why would you, unless you are a machinist or shop engineer. I know you have not spoken to a machinist, because I know what they will tell you, because I have asked, and purchased. It's usually about 'cost efficiency', not machine efficiency, unless the required part is holding up production, as in a steel plant. Achieving cost efficiency involves selecting a process flow that utilizes the least amount of resources, both capital and labor. I suppose that you could cut blanks with a water jet, throw the waste in a recycling bin, and go to the mill. You could also use a CNC to build them, completely. I'm just saying that for this simple round part, no one would do that.

"asked and purchased"? Ok - point made. You have no knowledge of the technology I'm taking about - fair enough

For what it's worth, I'm the guy that from machinists up, ask/discuss, the best way of doing something.

Just taking your band saw "I see ..." slag; So you tell me; what blade width and tpi should I use to cut a 5" circle in 1" plate? What blade speed for what alloy, be it any of the aluminium range, or any of the steels? How do I tell if a blade is cutting well, or poorly? What technique should I employ to follow a curve?

Do you know anything about what I'm talking about?

And there in lies the difference. I can actually do it - in my sleep. Whereas you have to ask. Now what is the 24 per shift, band saw guy likely to tell you? It sure ain't going to be "oh do this and you'll get 240 per shift" if it replaces him.

Typical billet will have the usual striations, and the first chuck setup can be a fiddle if the saw cut is not dead square. It may be a light first facing cut or a significantly intermittent one - so you have to allow for min two passes.

For the second cut, the blank has to be either on a face plate, or snugged to the chuck face (or jaw steps, if the chuck is accurate enough)

Now you can face, profile and bore the center (3 operations in 1 setup)

Sounds like you are assuming that the O.D. of the blank is at final diameter. Round aluminum bar from the mill is too rough, and varies too much (size, roundness, and surface finish) to be useful as-is for virtually any end-product.

Your "second cut" setup is what I do on the first setup: Slap the blank in a three-jaw, cut O.D., I.D., and profile. Facing (as a step all its own) is only necessary if the profiling doesn't include the entire surface. (If the O.D. can't be reached sufficiently due to the jaws, then it can wait till 2nd operation, holding the part by the I.D.)

Second setup is simply flipping the part around and using either soft-jaws or an expanding collet. Cut whatever couldn't be reached in first setup. Done.

What quantities are you expecting to produce? That will have a huge impact on manufacturing costs.

Will you ever need to re-profile or skim the wheels to compensate for wear? If so you need to consider the effect the heat generated will have on the epoxy. I have had this problem on much smaller HO/OO wheels where an attempt to re-profile them melted the insulating bushes.

Insulated wheel/axel sets are not required. This will be a composite wheel that combines strength, stability and low cost.

I wasn't suggesting you would need insulating bushes (although the epoxy will do that anyway) but that it might not be possible to re-profile worn wheels due to the heat damaging the epoxy.

OK. I think we have gotten a bit off track. This is a hobby project, not something that requires CNC, lasers or water cutting technology. The wheel is a pretty basis and simple turning job. I have designed fixtures to aid in my machining. I could have used a disc cut from plate or sliced off a round either from steel or aluminum. I don't feel for a model that I need the strength of steel. For my purposes, aluminum fits the bill. I, most importantly need to use only the tools I havemachine for this. My plan to use an epoxy/aluminum composite is based on the strength of epoxy (10000+psi compressive and adhesive tensile sheer of 2800psi, Devcon plastic steel liquid) being adequate to hold it's shape after cure and hard enough to be machined (Durometer 85D). In the real world, these wheels won't see loads anywhere near what a large scale live steamer will see. My way of doing this has been thought out very carefully taking into account all the variables that would be a problem for a full size locomotive. I won't see 50000lb loads and 40mph speeds of a full size locomotive. I'm pretty sure my design will work just fine for my application, but I thought I would toss it out for your thoughts. Being engineers, I think you may have applied full scale senarios to scale ones. Dimensions may scale down in linear ratios, but not so for weight or strength. You have to remember that I am doing this as a hobby and not to make money. Just the opposite, I want to make it as inexpensively as possible but employing good engineering principals, because that is the right way to design anything. I hope I haven't wasted anybodies time on this discussion, but your comments will always be welcome.

P.S. I could have presented this discussion on several other forums, but I felt CR4 was better equipped to handle it.

In response to all who think I should use a solid steel or aluminum piece of stock; I completely agree that solid material is the best way to go, but I'm thinking along the lines of "experimentation"; going where no one else has gone before. I might also add that; going the solid route, will take more time, produce more waste and use more electric energy. Epoxy (if it works) can be mixed and poured in less than 5 minutes. Curing time will take time, but that is not actual time spent. The machining shouldn't take more than 5 minutes to accomplish, even for an amateur machinist. Another thing I should point out is; I don't have access to a CNC machining center or water cutting machine. I only have an 11" Sheldon lathe and a Taiwanese vertical mill and a metal cutting bandsaw. I still think the cost per wheel will come out less overall.

As to strength of the tread, only an actual test will tell. I estimate the weight each wheel will carry to be around 10 pounds. The tread alone won't support this weight without deforming, but I think the epoxy with it's compressive strength of >10K PSI will oppose any potential deformation.

This is one of my pet projects and I will keep you informed with my progress. I have the resin on order; just waiting for it to arrive.

I still think the cost per wheel will come out to be less overall.