Comments on my new build plan?

[RodRico]

Guys,

After much consideration, I've decided to employ centrifugal investment casting to fabricate my engine components. I made this decision for a number of reasons: 1) it produces a surface finish that's very good (typically 125 microns); it facilitates fabrication of complex routing passages for oil, fuel, and air through the rotor; 3) it lends itself well to outsourcing during volume production of affordable components; and 4) it allows use of smaller, lighter, and more affordable tools in a smaller and more affordable workspace. I know investment casting takes skill, and I'll have to learn a lot before I'm successful, but the same would be true if I elected to attain clean finishes using a mill and lathe alone.

Investment casting requires molds. In some cases, a component model for mold making will be produced by working machinable casting wax using a DSLS 3000 precision micro-mill. Other component models will be 3D printed using the Liquid Crystal HR then cleaned up using the micro-mill. After the cast part has cooled, it will be cleaned up using the micro-mill, never taking off more than a couple of thousandths of either 4032 aluminum or Maraging 350 steel (which machines like 4130 steel in the annealed state). I believe the fact that I'm never taking off more than a couple of thousands of metal using the micro-mill will allow me to produce an accurate and finely finished surface.

Your feedback is, as always, valuable. Please let me know your thoughts.

Rod

 

[MikeHalloran]

If the USB Option is what I think it is, you should get it, because real parallel ports are disappearing from new computers right now.

Machining steel, even with very light cuts, will require cutting oil or water soluble coolant, either of which will make a mess in an open machine without even a hint of chip trays, like the one you propose. Further, the chips or the coolant or both will get into the stepping motors and the leadscrews.

You will probably have to spray lube oil on the ways regularly to keep them from sticking.

Machining steel, even with very light cuts, requires a lot of force, which will bend that aluminum table. Securing a decent vise to the aluminum table will deform the table, most likely permanently.

My first exposure to NC machining was milling precise circular parts from aluminum tooling plate, ca. 1970. Everything worked great, until the cutter hit a hard spot in the plate; the stepping motors would miss a step or several steps, and because the controls were open-loop, whatever geometry was generated after the missing steps would be off by that number of steps in that direction until the program ended. We produced a LOT of expensive scrap.

The machine you propose has encoders, which should detect missing steps, but ... then, what? Is there some part of the program that goes back and tries again, perhaps with extra current drive to the steppers, or does it just trigger an alarm, and stop? In commercial duty NC machines, steppers were replaced by hydraulically assisted steppers, then by full electric drives, and ballscrews replaced split nuts, and various forms of low-friction ways appeared. Hobby machines are sort of staring over, and will have to overcome the same problems, probably eventually converging toward the same solutions.

Since you're proposing minimal material removal, you would probably be money ahead to buy a real machine tool, even if it's been heavily used, even if the controls are completely obsolete and need to be replaced before you cut a chip.
OR, rent time on someone else's real machines,
OR, just contract the work out.

The machining time, and hence product cost, should be minimal if your blanks are near net shape. ... but not TOO near. ...

DO be careful that your blanks are sufficiently oversize to provide machining stock despite fixture uncertainty and part tolerances. The same outfit that provided the NC milling adventure also designed some parts to be made from sand castings, and the castings as designed didn't compensate for tolerances. So it the castings were just quickly tied down after indicating the major features, and the program run as written, we were machining 'secondary' hole features where there was no metal. After a few expensive pieces of scrap were produced that way, the remainder of the parts were done manually by experienced machinists, who verified that the casting was set up in such a way that there was actually metal available to cut wherever we wanted a hole. They probably would have been cheaper if cut from billet, even if that meant 95 pct stock removal.







Mike Halloran
Pembroke Pines, FL, USA

 

[RodRico]

Mike,

Good inputs, thanks!

I'm going to try and enclose the workspace near the mill chuck and do plan on using coolant. I also plan on measuring everything, so I should know if I'm having problems. My current life situation precludes me from having a large mill; the whole shop needs to be readily transported on a pickup truck. If it weren't for that, I was all set to buy a Tormach PCNC 770. If worse comes to worse and I can't attain the results I need, I'll have a professional machine shop do the critical cleanup work.

By the way, the folks at MicroProto say the DSLS 3000AB I'm planning is the same as the Taig 5019DSLS which has ball screws and NEMA 23 mounts with 200 oz-in steppers. I'm going to ask the seller whether it's possible to buy the mill with 425 oz-in NEMA 23 steppers. If you look at the Taig site, they do sell the same mill sans steppers, so I could go that route if need be. I'm planning to have my fixtures made by a professional shop, and I'm confident I could have them make a steel copy of the mill table.

It may also be worth noting that the largest steel part I need to clean up has a diameter of 2.900" and is 0.394" tall. It will lay flat very close to the table surface with no cantilevered holding fixture and a very short Z axis column. The largest aluminum part has diameter of 6.100" and is 2.195" thick and I will be drilling/threading small diameter holes in the flat surfaces.

Rod

 

[MikeHalloran]

I can't help but think that you're stitching your parachute after jumping out of the plane.
It can probably be done, but it's a big challenge, and the odds are against it.

A pickup can carry a small surface plate and a couple of toolboxes, so you can measure stuff as it arrives, but I think you'll be way ahead by farming out the machining entirely, as opposed to learning as you do it, and working with what I hate to characterize as toy machines just because of a lack of space.

Can your pickup pull a box trailer, or carry a camper shell?




Mike Halloran
Pembroke Pines, FL, USA

 

[SwinnyGG]

Tolerances, tolerances, tolerances.

All the same feedback from the last discussion of machining these parts still applies.

The machine you are looking at is not capable of producing the tolerances required to produce engine parts that will operate well.

Investment casting is capable of producing excellent surface finishes- IF the blanks are well made. Whatever flaws appear on your blanks will appear on your cast parts- you won't get 125 RA from your castings unless you are much better than 125 RA on your machined parts.

I understand you want to make everything yourself, but all the things I said in the last thread apply regardless of process.

In your current line of thinking, you are doing two things at once- developing an engine, and developing your own ability to manufacture.

These two tasks should not be done simultaneously.

 

[RodRico]

Mike,

I'm on the plane but haven't jumped out yet.

Do you have objections to using the "toy mill" to clean up 3D printed models? That's the main reason for it. If I find I can't get good results cleaning up the metal parts, I'll send that work out.

Rod

 

[EdStainless]

The micromill would be great for doing wax work.
The idea of printing and machining waxes is a fine development route.
Getting the size right will be a real trick, the wax-mold-metal shrinkage factors are a mess sometimes.
You will need to leave more than a few thou of stock in order to get good machined surfaces and proper metallurgy at the surfaces. Their are standards for investment casting and recommended min stock removal amoundts.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[RodRico]

Ed,

Thanks. I'll go look for those standards.

Rod

 

[enginesrus]

You need a bigger pickup truck.
The very minimum cnc mill would and should be a VF2 Haas size machine. It will have the strength and rigidity and accuracy to do what you need done. Anything else is a waste of time and money. Unless its a bigger and better machine.
Personally I prefer Fanuc controls, Haas are fairly good for the price though. The mini mills and hobby mills are okay if you are making pocket watch parts maybe. They have no power to cut and no rigidity, and is like tossing $ in a garbage can. Stepper is not the way to go for a mill.

 

[RodRico]

jgKRI and enginesrus,

Did you miss the part where I'm using the mill to machine soft models used to make investment casting molds? That's a significantly different application of these micro-mills than using them to make the parts out of pure metal stock.

Granted, I'll likely find I can't attain the tolerances I'd like using the micromill to clean up the cast parts. If that's the case, I will send the cast parts out to be worked.

Rod

 

[SwinnyGG]

Did you miss the part where I'm using the mill to machine soft models used to make investment casting molds? That's a significantly different application of these micro-mills than using them to make the parts out of pure metal stock.

I did not miss it.

You can certainly produce wax cores using that tool- because you don't need tight tolerance control. The challenge there will be tooling and workholding so that your cores are high enough quality. But that's an easier problem to solve.

You won't be able to finish your parts on that machine, which is what I meant when I said the same concerns as my last post related to this machine work were still applicable.

Mechanical repeatability of that machine is only .0005". And that's under ideal conditions. You need sub-tenth resolution so that you can produce parts to tenth accuracy. That machine isn't capable.

 

[kjoiner]

Have you considered 3D printing cores for investment casting?

https://www.stratasysdirect.com/app...terns/3d-printing-foundries-stereolithography

Kyle

 

[MikeHalloran]

I used to tell prospective employers that I could build an automobile, complete, starting by walking into an iron mine with a shovel.
... and that it would take just a little longer if I had to make the shovel first.

That's as true now as it ever was, including just a little hyperbole, but it's still just a hypothetical, and now I don't have enough time left on Earth to actually do it, should such a bizarre challenge actually appear.

... which is sort of where you're heading; you will run out of time.


I have observed, along the way to wherever the hell I am, that any project comprising more than one large unknown is absolutely doomed to fail. I can think of no theoretical reason why that should be true, but in my experience, it always works out that way.

You have stated an intent to develop an unusual engine and bring it into production.
That's a damn big unknown all by itself.

You have further demonstrated an intent to acquire the skills to produce every part of said engine by your own hand. That's the second big unknown.

Your latest meander suggests an intent to make many/all critical parts by investment casting, and of course to personally become a founder along the way. It will take you some time to get that right. ... but what if some of the highly stressed parts need to be forgings? Will you homebrew the process and equipment for that? It won't fit in a pickup truck, and it could take a lifetime to get it right.


If you are going to be an engine designer, concentrate on doing that well, produce some proper drawings, and leave the fabrication to people who can bring their aggregate experience to bear on just that.

You are starting to show a little skill at project management, for instance by starting with a simplified prototype. Develop that skill by using other people's fabrication skills, so you can keep going toward your goal without being distracted by learning skills that you can rent.

In software, I think project management with speed is called Rapid Development, or something like that. Basically, it means, fail fast, fail often, and keep going. So do that.


Mike Halloran
Pembroke Pines, FL, USA

 

[SwinnyGG]

I have observed, along the way to wherever the hell I am, that any project comprising more than one large unknown is absolutely doomed to fail.

This is my experience as well- I suspect it is due to the fact that when things fail, every new process or procedure is a potential root cause; doing two of them at the same time prevents you from optimizing either one through trial and error.

I've said it in my previous posts, and Mike is now saying it as well- concentrate on developing the engine first, if you every want your concept to see the light of day.

Once you have a concept that works (which means you have a prototype that not just runs but operates per design intent), then and only then should you focus on manufacturing.

 

[RodRico]

kjoiner,

Yes, I've considered outsourcing the whole job to a rapid prototype shop. I'm retired, and part of the what keeps me going every day is doing this work myself. I will certainly outsource if I fail and tire of trying.

Rod

 

[RodRico]

MikeHalloran and jgKRI,

I was in the R&D buisiness for the last 20 years of my career. I ran large teams comprised of engineers from every discipline in the development of my own inventions in the field of radar and electronic warfare. My personal contribution ranged from hands on engineering software and digital to project management and marketing to government labs for funding. We always started with far more than one unknown as well as several unknown unknowns, and not a single one of my creations was a total failure. Having only one unknown simply isn't R&D to me.

The technology and process on my engine project is the *least* risky of any project in my long history of development. The probability of attaining the desired performance is, however, about the same.

As I've said before, I do all this work myself because I'm retired and need something like this to keep me from going senile. That being said, if I simply can't do what I need to do to make this work, I'll send it out. I simply want to attempt doing it myself first so I'll have more fun and gain more hands on experience in the relevent knowledge and skills.

Rod

 

[SwinnyGG]

Neither of us is saying that you WILL ABSOLUTELY FAIL.

What we are saying is that your approach makes your development much more difficult. What you are planning to do is the exact opposite of the way mass market products are developed.

If you truly care about getting your product to market, you'll focus on developing the product. That isn't what you are doing.

 

[GregLocock]

FWIW I've seen a functional (smallish) steam engine made entirely using hand tools, from bar stock. Yes it is possible to make things using less than optimal tools. But as the Irish man said, if I wanted to get there, I wouldn't start from here.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[RodRico]

jgKRI,

I am very interested to see if my engine concept works and, if it does, I'm confident it will find a market. I have enough confidence in it to have spent $9,000 on patents. I am not doing this to make money, however. I'm doing it for fun. I have a lifelong love of learning new things. It's what drives me. While it's responsible for my successful career, it's not why I entered engineering; learning led to engineering, not the reverse. Designing my engine has provided untold hours of fun. Making it will add many, perhaps many many, more.

GregLocock,

Many people enjoy making small engines. In fact, there's a large community of them from around the world on the Home Model Engine Machinist website. Though many of the engines showcased there are steam engines, there are some wonderful internal combustion engines among the projects such as a V-12 Merlin, Harley Knucklehead, Straight Four, and so forth. YouTube is, of course, the showcase for many running models of internal combustion engines. There are too many to list, but here's a nice 12 Engine Compilation to get you started. All these findings are certainly what got me started thinking I could build my own engine using modest tools.

Rod

 

[GregLocock]

I built one when I should have been studying for my high school exams.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?