3d printing and casting

[chuckt]

I am thinking of undertaking a fairly major project. It's a 3d printer that can make foundry molds directly out of sand. I want to do that in support of hobbyists. I know this is being done out in the real world but it's expensive and I could not find anything that is on that level. I also make musical instruments and this looks like a way to increase my efficiency there as well. So I was curious to see if anyone here has an opinion about it.

I plan to sell plans and parts as well as do custom castings. I will have almost all of the work that goes into the machine available as licensed open source. I think that the open source 3d drafting software that I have been able to locate is not really ready for prime time so I may wind up using something that is not free for that. everything else appears to be available from the open source community.

If there is anyone else thinking about this I would be very interested to hear about it.

[RCW]

Please keep us posted on your progress.

[chuckt]

I have been doing some research. There is a machine that is open sourced called RepRap. It seems to have everything that I basically need already designed. Here is a link:
http://reprap.org/wiki/Main_Page

Since I have been getting into it, there seems to be quite a few academic institutions doing what I want to do. It is pretty experimental stuff. There is also plenty of free software available. I am excited. <<grin>>

[jpfalt]

The method you are talking about is an option on 3D Systems commercial equipment. For one or two off, it is worthwhile, but gets spendy fairly soon. It works by spreading layers of sand, powder and binder. The Rep-rap mentioned is a fusion deposition printer where plastic is melted and extruded through a small nozzle to build up a plastic part.

Another approach you may want to consider is use the Rep-rap type machine to print the pattern and then use the pattern to make green sand molds. The advantage is that for multiple castings, the unit cost is lower.

[chuckt]

So... Yes the reprap makes 3d positives from plastic. You could use it to make patterns. There are commercial units that make molds directly. They're really expensive. Here is an example:
http://www.youtube.com/watch?v=SgrGhJ5u5oM

I want to build a machine that does this, but cheaper. I am thinking of using sodium silicate as a binder instead of commercial epoxy binder. Here is a link to a machine that does some of what I want. The build surface moves up and down and the head moves in the X and Y axes. http://www.bitsfrombytes.com/usd/

Those darn things are $2-4K and they don't do what I want. I bet I can make a frame and the mechanics just like a RepMan and use the electronics from the RepRap. I would have to design the stage mechanism and the depositor head. Neither one of those looks very hard. I would also need to experiment with the sand and binder to get something that I can remove from the stage without breaking it. You can either bake SS or gas it. I would probably need to add wheat flour to the sand so I can get it off of the part after the pour.

Does that make sense? Or does it sound like an expensive folly?

[jpfalt]

The problem is that you really aren't going to be able to put a sand/binder mix through a print head. The only way to do what you want is to spread thin layers of the sand/binding powder and use the reprap head to deposit a liquid activator to set off the binder.

The 3D Systems unit spreads the sand/plaster mix with a screed and then uses an HP ink cartridge loaded with "clear ink" to set off the plaster, layer by layer. Sodium silicate binder requires CO2 gas to harden it, and I don't see an easy way to dispense it and maintain accurate casting surfaces in the mold you are printing.

There are enough methods available to make castings that you can probably find an easier way to reach your goal.

[Adirondack]

There is already a company doing this using a proprietary system. It is INCREDIBLY expensive to have something printed! That is why only heavyweight companies like Ford and NASA can afford to use them. When I had a 18" gauge truck project quoted from them it was over $8k just for a single set of sand molds, no metal pouring or shipping included.

It's a great idea, but I'm sorry to say it's not for home use. Not now, and probably not ever.

Sorry.

[chuckt]

You can check out what I have done so far here: http://chucktilbury.com.

Go to projects- > 3D printer. I intend to have it going by the end of this year.

[Pipescs]

I am thinking of undertaking a fairly major project. It's a 3d printer that can make foundry molds directly out of sand. I want to do that in support of hobbyists. I know this is being done out in the real world but it's expensive and I could not find anything that is on that level. I also make musical instruments and this looks like a way to increase my efficiency there as well. So I was curious to see if anyone here has an opinion about it

First off, Please do not take anything in this post as ment to be offensive.

Based on the quote above I do not believe you will find a market. Most of the Live Steam Modelers (Hobbiest like me) I have met, are looking for a product to be a raw casting they can machine, or a fully machined part they can bolt into thier model. I have yet to meet anyone looking for someone to ship them a mold that they would then have to take to a foundry.

You mention paying for a 8K for a sand mold. Even if this was a hundred dollar mold, this would be an extremely high risk to you as foundry sessions seldom come out perfect the first time around. When you pay a foundry such as Cattail to cast from a patten, they may end up playing around with the gating, risers, or any number of variables to get you a good part. If you send them a made up sand mold to do the casting, you will find yourself in a position of getting what you get. That is if they will deal with you at all.

Another approach you may want to consider is use the Rep-rap type machine to print the pattern and then use the pattern to make green sand molds. The advantage is that for multiple castings, the unit cost is lower

Also a person in the live steam industry, (that is a supplier), would be more interested in having completed patterns develped that they can take to a foundry to have molds made and castings produced. This is more inline with the quote above. I have developed a few patterns for people such as this. The ramming up of a green sand mold, at places such as Cattail, is in the range of less than 20.00 per mold . As said in the quote above, once the pattern cost is absorbed the molding cost is minimal.

As a modeler who has taken it to the point of building my own furncace I have done patterns in both RP and the traditional wood. I use a company called shapeways to produce my plastic RP patterns but have recently run into the issue of cost verses time. When a pattern reaches a certain size it becomes less cost effective to use RP based on material cost.

I will say I have enjoyed the detail you can put into a RP pattern and will continue to use it for detailed parts that are small.

As I say, this is just the rambling of a fellow modeler, and don't let anything I say keep you from your dream. I had quite a few "nay sayers" when I built my furnace, and started doing my own castings. Yes, a lot of the nay sayers were right, but I sure have a good time.

[chuckt]

First let me say "never stop telling me stuff". If you are telling me what you think then you cannot offend me.

There are a lot of naysayers. That's OK. I have never ever worked on a major project where there weren't loads of people telling me I was crazy. It certainly is possible that I am.

Basically, what I am doing is making a machine that can make metal molds that have the advantages of lost wax casting, such as inside cuts and the advantages of green sand, such as quick and inexpensive. Also, if someone wanted to put a plastic extruder on my machine instead of one for sand binders, that will be not only possible, but easy. I am shooting between the RepRap hobbyist machine and a professional quality 3D printer. If someone wants an all-up 3D printer, they could buy one of my machines. If they want to tinker they could build one from scratch using my open source drawings. Anything in between will work, too.

The advantage of have a 3D printer to make metal molds is that the pattern is in software. I can cast a pewter dragon on one heat and a bronze machine part on the next. I never have to store a physical pattern. Also, I will make 3D models available for all of the parts that make up the machine. There is lots of stuff that this machine is good for. I am willing to risk the effort and money in hopes that I will find a market. Time will tell it all.

(edit for clarity)

[scmods]

A couple of thoughts.

First, how will the sodium silicate be hardened in a timely manner without pressure or special atmosphere? Will the finished sand mold be strong enough to withstand shipping to a foundry or is it to be cast in house? Will the elements be porous enough to allow for gassing off during and after pouring? How about gates and risers? How about cores collapsing diring cooling to prevent tears during shrinkage?

There were two factories that made castings here years ago using sand. Cores were baked and puttied then covered with a wash for smoother surfaces. Components were molded, assembled, poured and shaken out in fairly quick sucession for small items, or molded, poured and cooled overnight for large items like machine frames and large rollers. On the large items, BC would complete the pour, dump the cupola, turn on the roof fans and leave with the molds afire from the oil burning out of the sand.

Second, although Stebbins Engineering makes a catylized polyester silica mortar for refractory purposes, it's intended use is for thin joints in refractory brick, not general construction purposes. The current state of epoxy products may indeed be sufficiently refractory to be utilized as a binder, but if it is, I'll bet it is expensive. If it isn't, there's more gas production to be dealt with during a bake or pour.

How about a UV catalyzed plastic like the dental products or the aforementioned nozzle build up to produce your mold. This would be used in a manner similar to the GM plant at Massena, which produced cylinder heads. They used a permanent mold to produce a foam casting that was then embedded in sand, gated and risered and left in place for pouring. The Styrene type castings were storable/shippable and the sand mold was made just prior to pouring so it was fresh. The Styrene was no match for the hot aluminum and disappeared into gas upon contact.

Your basic design is still digital and can be modified and reproduced at will. You could ship the design to a customer over the net, sell the plastic mold, or manufacture and sell the foam castings if it was not your desire to produce the metal castings yourself.

It is commendable to have an individual taking up a challenge such as this and I wish you the very best in your endeavour.

Have a good time and enjor the froits of your labor, so few do these days.

Bill Walck

[chuckt]

I recently found a product called HydroPerm made by USP. It's a type of plaster that is supposed to stand up to the temperatures of casting non-ferrous materials and it sets using water. I definately plan to include that in my experiments.

It may be possible to use sodium silicate with an aromatic esters such as Propylene Carbonate that cause it to harden. If they can be added to the sand ahead of time and then the SS is deposited, that may work. Or perhaps the SS can be added to the sand and the catylist deposited. One problem that I can see is that the SS is so thick that it does not soak into the sand very well. There may be a wetting agent that I can add that will help. Plain old dish soap is actually a powerful wetting agent, but it's got a lot of other stuff in it that may not be good.

The main idea is to make the castings myself. I am not really thinking about shipping molds anywhere. Also it is very easy to take an existing 3D model and create a mold. I will use something like Heeks CAD for that type of thing.