Geshwender PSRU Business

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
275
those are really nice castings.
Does the OP know what type patterns were used?
Were they metal patterns for shell molds?

If the patterns can be designed so that no loose cores are necessary, then direct per-unit cost of shell molding and casting will probably beat any process especially cnc.
OTOH the metal patterns will probably be made by cnc these days - and then loads of hand work.
(In the old daze, they were made by casting from wood/resin master patterns, then welded up to fix any defects, then tons of handwork.)

I'm an amateur that has made conventional matchplate patterns out of metal and resin, some wood (wood matchplate for one larger casting). & metal shell core boxes for shell cores; as well as worked with a half dozen mainly cast iron foundries, but also an aluminum foundry, to cast commercial products. The iron items were cast in batches of 10 or a dozen. For the aluminum product i made multi-impression matchplates (2-up), order multiple was 50 pcs. 3 foundries seriously offered me a job or to send me work based on pouring from my patterns. Most of the product i also had heat treated at one of the larger east coast heat treaters. A couple decades ago i studied the processes out of interest for making and selling niche items. I worked & studied to take cost out of the pattern-mould process. I have not had castings made since about 10 years ago when a cast iron foundry went bust and "lost" 2 sets of my patterns.

That industry continues to develop and move faster as well as the rest of technology. But they are in business to find and sell product.

smt
 
Last edited:

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
8,042
Location
Rocky Mountains
I'm an amateur that has made conventional matchplate patterns out of metal and resin, some wood
And IMHO the OP needs to find someone like you that can melt/pour enough metal to make them in batches of 10 or so. There are plenty of hobby level casters with the ability and equipment to make castings as shown and some could even do the heat treatment. If it is discovered there really is a need for more units than this kind of production can supply - then - the higher rate options should be considered.
 

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
275
I'm fascinated by pattern development, but my methods are old school hand work & about a dozen years out of date. I do know just about enough to be convinced of how little i really know. :)

AFA actual pouring & casting, i made commercial products and that is not conducive to cost effectiveness in a back-yard foundry. I've never tried it and am not interested. It is fun to watch pours at commercial operations, though.

If nothing but a form of self protection, commercial products should be made of known alloys to a given standard. For instance, in CI one might not care if they deal with a Mehanite foundry, but they probably should understand what class iron to specify and work with a foundry that is happy to use that method.

To mention some basic considerations that affect cost:
-work with a foundry that commonly/routinely pours the alloy you need.
-work with a foundry that likes to cast the physical size product you need. They will not have to work outside their convenience zone for existing equipment.
-work with a foundry where at least one of their furnaces and lines is sized to pour the amount of metal you can afford to batch your product with. A foundry wants to melt a furnace load of your alloy, and pour it all. So if they mostly have 950 lb and up furnaces & your raw snagged product weighs 20 lbs, they will want you to buy aproximately 25 of them at a time.
If they have a 150 lb melter, they are probably used to pouring small amounts of smaller items so might be ok pouring 4 of your 20 lb units as a batch. OTOH if your product requires a common alloy cast every day, they may batch your moulds on the line with a bunch of others for other customers whenever there are enough for a full melt.

Sprues. basins, risers, runners & other volume represented by the rigging can double the amount of metal actually used in the snagged item delivered to you. The rest goes back in the furnace for next time; or in some cases for a lesser product depending on the foundrys facility to maintain remelt alloys within spec.

At the pattern making level, develop patterns that suit your expected production level, balanced by the efficiency of the process for the foundry, x the quality level of the product you hope to sell. If you are only making one and it can't be cnc'd, loose wooden patterns & core box are appropriate, and you might figure on welding up defects because there's going to be a lot of other work for one-off at the machining level anyway.

A minor cost improvement at the one unit only/loose pattern/paste-up-core level is to mount the pattern on a matchplate.
This will allow no-bake moulding, though a core from split wooden boxes will be traditional and relatively costly. If quantities or repeat orders (even small orders) will be expected, then mount the patterns on matchplates, it will make a significant reduction in price over loose patterns,

If the casting requires loose cores, the next area to address is whether the product or patterns can be modified to eliminate cores. If not, the core making process can be a large cost factor. I actually don't know if cores can now be made by no-bake methods. At the time (30 years ago) that one of my products included a core; it cut the cost of a finished casting by some 2/3 to make a metal shell core box so they could be blown with resin bond sand. Perfect *cores, one every 90 seconds IIRC. Also, you could tell the foundry considered my job a nuisance at the hand-made core level. At the shell core box level they started treating me like a professional.

Next step, probably not applicable because of size, to the gearcases shown, is a multiple impression matchplate. If a ladle is tipped up over a mould, it is far less costly to keep it tipped and pouring if there are 4 castings in the mold ganged up on the matchplate and poured at once, than to step over and pour 4 castings one at a time.

Based on advice from one foundry, (for more cost reduction) I was exploring boxed patterns at the time the last foundry went under but have no experience. The types of patterns different foundries want to work with can be specific to them.

It you plan to produce several hundred castings per year on a steady basis (or more, way more) you probably can't afford *not* to develop metal patterns & use shell mold processes. (If suitable to the casting shape).

As stated, i have no idea or experience with modern lost foam processes. They seem to be the answer to complex castings. Many of which couldn't possibly be cnc'd.

*cores - the person who posted about die casting may have other experience. However, traditionally, among other reasons it is not broadly applicable to structural aluminum parts, is not only suitable alloys & hot short tears in complex shapes without cores as the casting cools, but the difficulty of coring. Shell cores are close to ideal because they are blown into a heated metal (split) core box. Resin in the sand causes it to cure where it contacts the hot metal. The process is timed (in seconds) for a wall thickness based on the foundry's experience with similar parts. The rest of the sand is dumped out for continued use, and the "shell" is removed for use. (Loosely analogous to slip casting ceramics) It can be seen that the amount of crush can be tailored, so the core collapses at a rate suitable to the shrinking metal as it solidifies. So the core does not cause distortion or tearing in the hot-short metal.

(Traditional molasses pasted up cores & similar bond sand cores accomplish this by easily crushing, but they are delicate, & difficult to make far ahead of time and stored for use. They generally require being made solid in 2 halves, baked, & stuck together. Many require wire armatures or metal chaps for support, where shell cores have enough strength & the prints can be designed integrally to resist flotation, torquing or other stresses, without metal re-inforcement. All hand work, and all essentially time consuming. )

AFAIU, lost foam can dispense with separate cores in some (many?) cases, but someone with direct experience should comment on that.

smt
 
Last edited:

Lendo

Well-Known Member
Joined
Feb 6, 2013
Messages
875
Location
Brisbane
As I'm not familiar with this PSRU I would be interested in seeing the internals if anyone has a photo or drawing, looks to be good workmanship.
George
 

J Galt

Well-Known Member
Lifetime Supporter
Joined
Nov 29, 2019
Messages
65
Location
KMAN, Idaho
Justin:

Is it possible that your V-8 conversion engine supplier could be ressurrected?

Leo A.
Leo,
I doubt it buy I'll ask. Both the principals are well into retirement I believe, one is for sure.

Regarding the casting cost, I would personally say at a minimum with a $15k selling price you're going to need to see $7.5k COGS or less. Amortizing $40k in molds into 12 sets of castings doesn't get you a good start and I would think machining the castings and then the gears themselves would be big cost drivers. If it was me I would think hard to ascertain if I could sell at least 40 of them at a minimum or find a way to bring the casting capital investment down. As has been mentioned, you can do a lot of machining on aluminum billet for well less than the $40k you'd be into just for tooling.

I hope that helps, again it's just my own opinions and estimates,
Justin
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
8,042
Location
Rocky Mountains
Amortizing $40k in molds into 12 sets of castings
This is the first number to pin down. How many units are going to be needed and in what time frame. Without that number the rest is just guessing............but then that number is also a guess without some good market research or an existing demand.

12 sets of castings is well into the wood flask, hand scratched runner and gate casting method range.
 

rv6ejguy

Well-Known Member
HBA Supporter
Joined
Jun 26, 2012
Messages
4,707
Location
Calgary, Alberta, Canada
There are lots of tire kickers out there. Unless you get non-refundable deposits, you may find that many of the 40 people saying they'd buy a drive for X amount will be nowhere to be found when the time comes to pony up the full amount. Just my experience over 25 years in designing and making specialty parts for Experimental aviation.

Your chances of getting 40 NR deposits on something which isn't in production yet at this $15K price point in this market is essentially zero IMHO.
 

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
275
This is the first number to pin down. How many units are going to be needed and in what time frame. Without that number the rest is just guessing
This^^^^^^
That's a succinct way of putting what i've been trying to rebut to the big dollar quotes by long winded details of the pattern making process. The wonderful thing about online cost projections is that none of the projectors have to specify what the money will cover. Process? materials? Nothing.

Someone has to define process & research the vendors.
It's a back and forth iteration: Find the mostly likely foundries, find the methods they use to produce the quality and quantity level you want to afford, then define your tooling around that. When you know what kind of tooling you should be using, then get the tooling quoted if you don't want to make it yourself.

I'm fairly convinced most of the numbers thrown online are by people who have never had to do binding estimates and sit down with a calculator and time sheets, materials costs , etc and spend a few hours defining the project.

Get a list of aluminum foundries within. say, 250 mi radius. Or a list of all the aluminum foundries in your state. Pick a few whose online presence clearly indicates client friendly, especially if they offer tours, and go tour them and start some discussions. Your objective is a fast education. You will begin to sense what types of tooling suit the type of casting at the level you can afford to buy them. Wood/resin patterns could be ideal to start so long as they are on matchplates, and should be suitable to no-bake or air set moulds if the foundry likes that process. There are some great green sand foundries in PA that do fine work off that type tooling, and that's only one state.

The Geshwender castings some have posted, though, look to my eye like they probably came out of no-bake or air set moulds due to the finer surgace texture. Wood/resin patterns work with that process as well. If the originals were made out of shell moulds, (requires perfect metal patterns) or if the OP intends to go first class regardless projected quantities & expects turn-key results with no personal input except check-writing, the high numbers bandied on here could be correct. If the OP can do his own leg work and informed project & process management, it sounds a bit rich for low quantity processes that have worked for a century, were refined to a high level during WW2, and have developed even further in the last couple decades.

smt
 
Last edited:

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
8,042
Location
Rocky Mountains
look to my eye like they probably came out of no-bake or air set moulds due to the finer surgace texture.
Could be the way they were done. My first guess was Petrobond: based on surface finish and the number that have been produced over the years.
The old pasterns would be a valuable asset, even if they needed some repair.
 

rv6ejguy

Well-Known Member
HBA Supporter
Joined
Jun 26, 2012
Messages
4,707
Location
Calgary, Alberta, Canada
I can't see how castings make much sense financially to test the waters of the market size here. Get the billet onto a big 3 or 4 axis mill and start carving. No shrinkage issues, less setup time, no heat treating and distortion worries. No big outlay past the billet.

Turn one out, build it, run it on an engine on a test stand, take photos and make a video for marketing. The casting process will soak up a ton of money whether you produce 1 or 40 sets of parts. Do a lot of castings to get the unit price down and you may be stuck with a lot of bits if they don't sell well.
 
Last edited:

rv7charlie

Well-Known Member
Lifetime Supporter
Joined
Nov 17, 2014
Messages
2,169
Location
Pocahontas MS
Not to mention the issue of voids in the casting; according to a friend who's done some aluminum casting, it ain't easy to get a strong enough aluminum casting without a *lot* of extra material (weight).
 

lganderz

Member
Joined
Feb 10, 2021
Messages
13
Thanks all: I really appreciate the feedback.
I can't see how castings make much sense financially to test the waters of the market size here. Get the billet onto a big 3 or 4 axis mill and start carving. No shrinkage issues, less setup time, no heat treating and distortion worries. No big outlay past the billet.

Turn one out, build it, run it on an engine on a test stand, take photos and make a video for marketing. The casting process will soak up a ton of money whether you produce 1 or 40 sets of parts.
Thanks all for the great feedback. Looks like my next move is to get a 3-D model that I can shop to CNC houses to get some estimates.

Leo A.
 

lganderz

Member
Joined
Feb 10, 2021
Messages
13
Leo,
I doubt it buy I'll ask. Both the principals are well into retirement I believe, one is for sure.
Justin
I'd appreciate the opportunity to talk to one of them if possible. I'm sure they have some nuggets of wisdom that would be worth the effort.

Leo A.
 

OldRVFlyer

New Member
Joined
Nov 3, 2020
Messages
3
Location
Northern California
I've been messing with experimental aircraft for over 55 years but have no casting knowledge. My question is: Are there reputable foundries out there that would be willing to take on a small profit potential casting run that involves the risk / liability associated with "aviation"? "Aviation liability" issues have doomed many well-intentioned projects over the years.
 

Hot Wings

Grumpy Cynic
HBA Supporter
Log Member
Joined
Nov 14, 2009
Messages
8,042
Location
Rocky Mountains
I can't see how castings make much sense financially to test the waters of the market size here.
It doesn't...........IMHO
For development CNC makes a lot more sense. I've presumed that this particular product had already been developed and is ready for or was in production at one time?
The question remaining is how big is the market. If you want to 'test the waters' by actually offering the product for sale a CNC billet would be a wise choice.

If the company had a lot of market data, or 'a good feeling' about the market size to gamble with, then straight to the best method, what ever it is, should produce the best profit or the most direct route to bankruptcy.

As far as voids, bifilms, hydrogen and other inclusions in the aluminum castings? Just like any other manufacturing process you find a set of parameters that produce an acceptable part and then control those parameters to get consistent results. Lots of parameters in aluminum casting.............:eek:
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
15,868
Location
Memphis, TN
Nowadays, CNC is probably cheaper for small runs. I know of one company who keeps adding CNC versions of their parts, because small batches of castings have gotten too expensive. You will need the brother in law connection for good small run castings.
 

Lendo

Well-Known Member
Joined
Feb 6, 2013
Messages
875
Location
Brisbane
Another problem with sand castings, is that to get them to run into all areas, they often need to be thicker and therefore heavier they they need to be.
I would have to agree with all suggestions posted here.
Do you think the casing design is optimum or could it be improved upon. I've worked in a Foundry and had parts made for Mazda Bell housing, but have been impressed with later designs, probably as a result of Computer Analysis placing material where it's needed the most. Just a suggestion.
George
 

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
275
My question is: Are there reputable foundries out there that would be willing to take on a small profit potential casting run that involves the risk / liability associated with "aviation"?
The airboat market is far bigger.
There are multiple competing gearboxes for similar engines as converted for AC.
Originally cast case, most new market entrants seem to be cnc'ing. Similar price range regardless case type: $3,200 - $4,500 or so for bolt-on reduction unit.
Same apparent supply problems as AC types for some vendors. :)
There also seems to be interest/potential demand for an analog to the belt drive box. Due to maintaining engine rotational direction for pusher props.
Perhaps one of the 1/2 dozen vendors would sell empty cases? Stinger/Torque/Rotator/Ox Box/Ballistic + a couple belt drives.
Perhaps the Geshwender chain drive would have a niche for the airboat customers wishing for a 3 gear, or a better belt drive, to maintain rotational direction?
Airboat boxes don't accommodate constant speed props; but otherwise solve the same problems faced by aircraft PSRU's.

Good luck!
 

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
275
Another problem with sand castings, is that to get them to run into all areas, they often need to be thicker and therefore heavier they they need to be.
Actually, this is backwards.
Some types of No-bake, and certainly lost foam (extensions of sand casting) as well as many actual green or petro-bond sand castings are used to make thin walls in sections that can't possibly be reached or machined by cnc.
The type cases under discussion, though, can be designed to avoid that, so it's down to cost.
Effective (strong, lightweight, thermally informed) engine blocks are a different story.
 
Top