HP just announced a new Metaljet 3D printer that prints in stainless steel instead of plastic. They say it will be able to do other metals in the future. It's quite expensive now (I think $50,000) but give it a few years, and like anything else in the computer world the price will come down dramatically.
$50k would be cheap. An HP MetalJet is $400k. I am trying to get one for work. One of the two 3D printers we have at work (the bigger one that I mentioned), was $250k.
Yikes, I figured Stratasys was a top of the line machine but I never figured $250K, OUCH! You could buy a whole lot of CNC equipment for that kind of scratch. It is no wonder that they are all over the Detroit area hocking their wares to suppliers and OEM's. Definitely out of reach for the average hot rodder. Cheaper ones are available.
If a person doesn't have a printer big enough to print their part, there is a solution. The trick we have used in places I have worked is to cut the CAD model into two or three pieces. We usually make a "Z" scarf in the model. We then print the separate pieces and the carefully clean up the scarfed ends and glue the printed model together.
I interviewed for an engineering position at HP when they were developing the metal 3D printers, which are supposed to be leaps and bounds faster than competition. They grilled me with academic/intellectual type engineering questions for about 2 hours (you know, the types you only see in textbooks that are the exception to the rule every time).....I had been out of school for several years, hadn't heard or used any of that information since, and totally botched that portion of the interview LOL. Needless to say, I didn't get the job. If you only needed one part, I believe 3D printed metal is a good option that will get better/more affordable as technology progresses. However, to make your water crossover in short production runs, I would think a more cost effective method would be a CNC router. If I still had access to a CAD program and wanted to make a short production run, this is how I would do it anyway. I have some foundry experience and a fair amount of CAD under my belt, but it's been a few years on both disciplines. I currently bathe in toxic chemicals for a profession. Aren't life choices awesome? Anyway, the idea is to split your CAD part in half, make a female part of the top half, a female part of the bottom half, and top and bottom (female) parts of the water passage in similar fashion. Someone half decent in CAD could knock this out in an hour or so from your existing model. For everyone's sanity sake, I won't go into full detail, but something along these lines (if you're having the college students do this for free, they could follow this list): In the CAD model of your part, highlight the whole thing and increase the volume to 102% (to account for shrinkage). Next, , put a horizontal plane down the longitudinal center (hot dog bun) of the part model. On each side of the plane, put an additional plane, say 3" above it and 3" below it - enough so the top and bottom planes are above and below the surface of your part. On the top plane, sketch a rectangle that has a perimeter a couple inches wider and a couple inches longer than your part. On the plane that's cutting your part in half, sketch a rectangle those same dimensions by superimposing the rectangle from the plane above. Now do an extrusion from the top plane to the plane going through the center of your part. This is something I don't remember too well, but somehow you do a shell of your part or something in a sketch along those lines so the extrusion function doesn't consume your part, but only goes to the surface of the part. What you're left with in the model is a rectangular block that goes from the top plane to the surface of your part/middle plane. Save this part model (the block only) as something different than your model, like h20xover_cope. Do the same thing for the bottom plane, extruding from bottom to middle plane/shell/surface of part and save that part model as an additional part like h20xover_drag. In a similar fashion, create extrusions of the water cavity surface to the top plane and bottom plane, saving these as additional parts h20_cavity_corebox_top and h20_cavity_corebox_bottom. Now open an assembly model, and insert your h20_cavity_corebox_top and h20_cavity_corebox_bottom. Mate the two together. On the thermostat inlet and the outlet ports on each side of the assembly, create small extrusions outward, or nipples if you will (core prints), then save the parts so the cavity top and bottom are both updated with the nipples (core prints). Create another assembly model, insert the cope model and corebox top, then mate them together using the top plane (put the corebox top inside the cope model if that makes sense). Create an extrusion in the cope model to match the core prints in the corebox top, then save the cope model with the updated core prints. In the same assembly, insert the drag and corebox bottom, then mate them together and repeat the extrusions and saving model with updated core prints. Go back into the corebox top/bottom assembly model and create 3/8" diameter extruded cuts 1/2" deep into the mating surfaces of both top and bottom parts on all 4 corners (dowel locating pins), then save. NOW.......FINALLY..... You have 4 part models. Have each one of those CNC routed out of a rectangular block of wood that is ~ the same size as the extruded rectangles from the parts models. From the cope and drag, you can coat each with mold release and create the top and bottom of your pattern using foundry mold (bondo). Screw these to your pattern board (countersunk screws), then bondo the screw-holes smooth.. Add some gates and risers, etc. with wooden blocks and more foundry mold/bondo, and now you have a pattern that is usable. If it gets damaged, you can separate the bondo'd molds and repair by repouring new mold into your CND'd parts. Take your corebox top and bottom and your pattern board to a foundry (or do this part yourself if you're equipped to do so). They will glue up sand cores in the core box, then sand pack a cope and drag around your pattern board. After that, they separate the cope and drag that are sandwiching your pattern, drop in a core, and pour the part. Once broken out of the mold, you have a part that is ready to be machined. It sounds like a lot of work (and it is), but if I understand correctly, you have college students doing the CAD modeling? You'd potentially only have to pay for the CNC routing if pouring yourself... ^Cores (which were made in a core box) in a mold ready to be poured^
there's a big difference between computing power and precision machinery, as far as the price going down over time. This stuff will never be cheap.
LOL, I have been trying to wrap my head around the above post myself. I need to sit down and read it carefully and decipher it.
If you've done CAD work before, my post will make sense. If not, my post will be the last couple straps cinching down my straight jacket as you all haul me off to the looney farm. Sent from my iPad using H.A.M.B.
Additionally, you could model the part to make a male top and bottom of the pattern instead of a female, which would eliminate the need for you to pour foundry mold resin/bondo to make your own pattern, but if you do it that way and there's any damage to your pattern during the manufacturing process, you're kinda hosed as you'd have to have the pattern CNC routed out of solid wood again. If you make female patterns like I described, the worst thing that happens is your resin/bondo pattern gets messed up, so you mix up a batch and pour another one to repair it. Sent from my iPad using H.A.M.B.
Yes, the learning curve is rather vertical with Catia. Funny thing, I design casting and forgings for driveline use here at work. However all I do is make the final product model and drawing. All of our prototype work is done in India or China. They are responsible for the models and drawings for the raw (un-machined) pieces.
Think of it as you're making two boxes both the same size that each split down the middle. Open up the first box and your part is inside. Open the second box and the water passage is inside. The first box is the mold for your pattern halves. The second box is the core box used to make the core. Sent from my iPad using H.A.M.B.
To me the obvious downscalable process would be either 3D-printed master > reusable form > wax > form > part or 3D-printed form > wax > form > part, depending on the envisaged rate of production. In both cases the 3D printing kit is fairly basic.
I yanked it from a google search. I thought it was a good example of a core with easy to see core prints and how they integrate into the mold. https://www.poclain-technicast.com/know-how/complex-parts/
How would you get to the wax part? I understand everything else you said, but I'm not understanding how it would be a repeatable process.
Good post Tim. Description isn't all that much different than doing it by hand. CAD just replaces a lot of hand-measuring , scaling, etc. Wish I was conversant in CAD, & basic computer. Marcus...
That's what I was getting at with cnc routing the sand molds. I don't have any experience with lost wax castings, although I am familiar enough to know how it works... I wonder what the pros/cons would be of lost wax vs sand casting with cores, etc. To me it seems like the consumable wax molds add a step, but maybe it's easier to cast several at once by gluing them to a tree? That's one advantage I see. Sent from my iPhone using H.A.M.B.
The Jalopy Journal
