Casting vs Fabrication - Small Tank Tracks

[Avadon]

Hello fellow metal workers,

I have some questions regarding making aluminum tank tracks for a robotics project. The tank tracks are 4.25" wide x 5/8" tall (length TBD) they are made up similar to the way an Abrams tank track is put together. I found a metal works company who specializes in turning small 6061 aluminum - rubber coated wheels and though expensive they can produce what I'm looking for. So now I'm back to the tank tread portion which I'll probably make myself. My experience: 10+years experience with TIG welding/fabrication and can weld most things. Many years ago I read a book on backyard metal casting because I wanted to cast aluminum tank treads. Back then the guy was using a Pyramid furnace (the company is no longer in business) to cast aluminum tracks. So here I am wanting to pioneer this venture again with many questions. After reading the beginner pinned thread and all the safety precautions, Aluminum casting seems a little scary/intimidating as well as an enormous investment of time to learn a whole new universe of skills and the safety surrounding them. Therefore I've come to these questions which I hope others can help me answer.

A:) Would a furnace produce acceptable enough results in fit and finish for this type of application? (or am I better off in fabrication?)

B:) How much cost investment is there in buying a high quality furnace, crucible and safety gear?

C:) Roughly how much time investment is there in learning to cast small aluminum parts?

D:) What is the mold material needed?

E:) Are there any up to date extensive youtube videos or good book recommends specifically for casting aluminum.

I have concerns with doing it safely as well making sure the mold is built correctly. Additional massive concerns about what the results will be after I Invest huge amounts of time and money in it. Will the results be good enough for this use? And more importantly will they be better than a machined/welded/fabricated alternative. In other words: Which route is shorter and produces the best results? I'll attach some drawing and pics of what I'm trying to manufacture.
 

Thank you,
Av

[Hay River]

Wow. You have your work cut out for you.  I can't speak to the casting directly but I have a few points that may help you decide. 

The castings will probably need a machining operation.  I don't think you will be successful casting in the cross holes accurately. 

Also what is the size of this part? It may save material costs by fabbing together.  Is this a contract job or your own government work? In other words are you working for free or do you need to bill it? The material savings you may realize could be consumed by labor in fabrication  if you have to bill the hours.

Cool project though.  Keep the pics coming.  

Bestof luck. 

[JNewman]

Are the treads you are looking at casting as small as the ones you show held by hand?  Those look to be die cast which is probably the best bet for something that small and detailed  if you were making thousands of parts but is way outside the scope of  backyard casting.  You could investment cast them making a mould for your wax patterns. 

If they are bigger than those small ones they could be done by sand casting which is more suitable to a backyard project.  You need a lot of castings for this project so I would suggest making a matchplate with multiple patterns on it so your moulding will take less time and you end up with several parts every time you make a mould.

With sand casting you are going to need to machine the parts as Hay River states.  Die cast and investment cast  you MIGHT get away without but machined parts are going to operate much more smoothly and die cast parts tend to be weaker. 

[CMS3900]

I would look into having them 3d printed in Aluminum, Steel, Brass ect.. The flash on them will probably less substantial than home casting, and probably cheaper. Also it looks like you already have a cad drawing of the part.

[DSW]

Personal or school project? I ask because if it's for a school robotics program you may be able to find a local foundry who might cast your parts with "excess" material from other pours at little or no cost. That's assuming you can deal with the time issue waiting until they have enough to get you all your parts cast. That would save you a lot of the hassle. You just make the master pattern and they do what they do best.

[Avadon]

I think if the castings needs machining then I should just have them all machined. In the last picture: I can make the pins, I already have the outer links, and I can even make the inner horns. What I can't easily make is the track pads. They are ~2.5" x 1.25" and they have holes in their lengths to accept the pins and obviously I need a ton of them which is why I was wondering what the best way to do this would be.

[JNewman]

6 hours ago, CMS3900 said:

I would look into having them 3d printed in Aluminum, Steel, Brass ect.. The flash on them will probably less substantial than home casting, and probably cheaper. Also it looks like you already have a cad drawing of the part.

Metal 3d printing is a very expensive process.  3D printers that will print metal  are close to a million dollars print very slowly and use expensive consumables. 

They are good for parts that are difficult or impossible to machine.    

[CMS3900]

1 hour ago, JNewman said:

Metal 3d printing is a very expensive process.  3D printers that will print metal  are close to a million dollars print very slowly and use expensive consumables. 

They are good for parts that are difficult or impossible to machine.    

Since were not supposed to post links here apparently I won't, but google 3d printing service and there is one with Shape in their name. They are pretty affordable and I know many people who have used them. For what they charge it beats boiling down aluminum cans in the back yard with questionable results. What I didn't catch the first time was how big the parts are that he is talking about making, roughly 2.5" x 1.25" x 5/8" The model tank pictures were throwing me off. I still don't think casting is a efficient strategy for his part.

Avadon, can you post up a picture of exactly what your looking to make, with dimensions? It sounds like a cool project and reminds me of the FIRST Tech Challenge robots I have helped the local league with.

 

[JNewman]

I went to the suggested website.  While $65 for a 1 off aluminum item that is 23grams may not be that expensive as there are no tooling costs.  But it is awfully expensive for something that dozens are needed.  The shown parts could likely be CNC machined for much less than they can be 3d printed if more than a couple were being made.   A solid block of aluminum with the dimensions you described would be about 89 grams so would likely be in the $250 each range going by the prices on that website.  . 

I never suggested melting pop cans.  The original question was about buying a furnace,crucible and other tools so I would assume the OP would be buying new material or appropriate scrap.  Another option would be taking a matchplate to a foundry and having them cast the parts.  However for only one set of tracks machining them from solid may be the cheapest way to go. 

[Smoggy]

Why are you making them out of aluminium?

[Avadon]

23 hours ago, CMS3900 said:

Since were not supposed to post links here apparently I won't, but google 3d printing service and there is one with Shape in their name. They are pretty affordable and I know many people who have used them. For what they charge it beats boiling down aluminum cans in the back yard with questionable results. What I didn't catch the first time was how big the parts are that he is talking about making, roughly 2.5" x 1.25" x 5/8" The model tank pictures were throwing me off. I still don't think casting is a efficient strategy for his part.

Avadon, can you post up a picture of exactly what your looking to make, with dimensions? It sounds like a cool project and reminds me of the FIRST Tech Challenge robots I have helped the local league with.

 

I'm making something similar to this only magnitudes way more advanced. It's kind of out of topic for the thread, but I'm making a deep learning android w/ strong AI.

It looks not so similar to this but you'll get the idea.

 

4 hours ago, JNewman said:

I went to the suggested website.  While $65 for a 1 off aluminum item that is 23grams may not be that expensive as there are no tooling costs.  But it is awfully expensive for something that dozens are needed.  The shown parts could likely be CNC machined for much less than they can be 3d printed if more than a couple were being made.   A solid block of aluminum with the dimensions you described would be about 89 grams so would likely be in the $250 each range going by the prices on that website.  . 

I never suggested melting pop cans.  The original question was about buying a furnace,crucible and other tools so I would assume the OP would be buying new material or appropriate scrap.  Another option would be taking a matchplate to a foundry and having them cast the parts.  However for only one set of tracks machining them from solid may be the cheapest way to go. 

Yes I'm starting to believe you're right. Machining a 1-off copy is most likely way cheaper than getting involved in my own foundry which sort of takes me away from the direction of robotics and into the field of learning metal casting. I'm thinking someone locally can probably CNC my track pad parts for cheaper than my investment in setting up all the necessary equipment to pour these pads in a manner that will produce impressive results. And of course no pop cans. I was only thinking of the best possible aluminum scrap.

1 hour ago, Smoggy said:

Why are you making them out of aluminium?

That's a good question. I thought of doing them in steel but the machining will be far more difficult and the weight will be dramatically increased. I think aluminum is really the best bet. I'm not going to be climbing any mountains with the robot so I think the aluminum should be more than durable.

[Smoggy]

Then can I suggest you contemplate using zinc, it can be just as durable as aluminium, little difference in weight and easier still to both cast and machine. Zinc having a lower melting temp than Al can be handled in a simple (but strong steel crucibe) Much of what you think may be cast from aluminium wiil either be aluminium alloy or zinc/zinc alloy. I suggest you also consider investment casting, lost wax system. A cheap rubberized mold can be use to cast many wax copies, which can then be attached together on a main wax sprue and coated in a refractory slip, this is then baked and the wax melted out leaving the refractory mold ready to accept the pour. thus casting many parts in one pour. The majority of jewelry is made via this method. There are many youtube vids demonstrating this technique. The above can be achieved with very basic and reasonably priced equipment, but do take time to study the subject and do not skimp on safety. No dampness near any stage of the process, molten metal and water spells potential major dissaster, wear sutable PPE and be aware of metal fumes, especially Zinc, although if working at the correct temperature it should not be a problem melting zinc,burning it is the main hazard, but always a well ventilated area. If I've not managed to put you off casting, study it further and if you still think it's a good idea ..... just take care!

The rubber treads, have a look around for premolded rubber sections there are many shapes and sizes available off the shelf usually intended for seals, such as the bottom of roll down garage doors, fridge door seals, glazing seals etc, it's not inconceavable that there could be just the right section to cut to length and fit nicely into a track section with a sutably adjusted design! Alternatly could a section of appropriately sized rubber tubing possibly be slipped over the track section?

[Charles R. Stevens]

Rubber tracks, either fabricated from baler belting like a snow cat, reperposed snow mobile rear track, or fabricated from a tire with the side wall cut off may be better. 

[JNewman]

25 minutes ago, Charles R. Stevens said:

Rubber tracks, either fabricated from baler belting like a snow cat, reperposed snow mobile rear track, or fabricated from a tire with the side wall cut off may be better. 

I like Charles' idea.   The grouser bars on a Nodwell flextrack would be fairly easy to hot forge under a press with a die.  The profile on a Thiokol or Pisten Bully grouser would be tough to make but an angle iron piece would work as well.  You would have to bend and weld the guide pieces that keep the track centered on the bogies.

The only problem with a snow cat type track is that your first  contact on the ground is steel or aluminum which does not give you as much traction on paved areas without carbide studs.  Off road they are hard to beat. 

[ThomasPowers]

My local scrapyard has some full sized steel ones in the pile....I would have suggested taking this question over to backyardmetalcasting.com and the forums there on a site that specializes in casting.

[Avadon]

I found this recently which gives a better show of how M60 tracks are made.

http://www.safestchina.com/products_image/m60-battle-tank-shoe-9322f.jpg

 

I think the easiest thing is instead of machining aluminum tracks and pouring rubber into them, make the whole track pad of rubber with either tubes or pins poured into them. I've seen this done with good results. It cuts down a very big step of having to either machine or cast metal and then pour rubber into it. If we are dealing with 20 tons then you definitely have to have the metal in there, but seeing as our robbie is under 200LB I think we'll be fine with just very strong truck tread shore 70-80 hardness treads.

[Avadon]

For anyone interested, there's a ton more stuff in small scale

http://www.aliexpress.com/promotion/promotion_tire-tracks-promotion.html

 

Our robot is near 1/5th scale.