# RR Merlin Build

### Help Support Homebuilt Aircraft & Kit Plane Forum:

#### Jay Kempf

##### Curmudgeon in Training (CIT)
Creating an accurate digital engineering model is the first step in making something like this possible. It could be used to create the tooling and processes and it wouldn't require an enormous amount of resources other than digital and time ones. And you could sell copies!
What is the status of Merlin engines out there? If someone were to engineer a replacement would there be a world market for restoration or museum projects? The killer is that if you wanted to use a fully engineered engine to replace a Merlin you would probably have to pay for all the development work shared amongst the few who play and that isn't cost effective. But wouldn't that be a milestone, that there was a market to resurrect designs that are that far back in the industrial revolution timeline. No one has done the kind of development on government dollars in augmented piston engines since that era just because there hasn't been a need. The dual inline V8 (making V16) seems like the best alternative I have seen recently. That still has merit in terms of off the shelf parts but the work to make an engine work in the 2700RPM range is significant. We have lots of auto engines that highway cruise with no power in that range but none that actually make any significant power there.

I wonder what engine would be developed these days if someone had the task of designing a P51 from scratch? Today's engines tend toward higher revving boosted engines for light weight and power. With the requirements of a prop driven high altitude long range machine and the limitation of piston engines I wonder what would have been developed. At the end of the development period during the war it seems that there was still a big disagreement amongst the competitors whether a big radial or an inline V watercooled was the best way to skin the cat.... Hmmmmmm?

#### Autodidact

##### Well-Known Member
I'm not sure there is a paying market for a new Merlin, yet. But the engineering design is already done and there is at present a lot of practical knowlege of these engines (and others); there are currently airframes pretty much being built from scratch and at some point someone will be willing to give making new engines a try (TVAL have already done this on an only slightly smaller scale). And if a detailed digital model existed then it would be a big obstacle out of the way. Casting pattern designs could be generated, tool paths for the milling, etc.

#### Topaz

##### Super Moderator
Staff member
Log Member
...I wonder what engine would be developed these days if someone had the task of designing a P51 from scratch? Today's engines tend toward higher revving boosted engines for light weight and power. With the requirements of a prop driven high altitude long range machine and the limitation of piston engines I wonder what would have been developed. At the end of the development period during the war it seems that there was still a big disagreement amongst the competitors whether a big radial or an inline V watercooled was the best way to skin the cat.... Hmmmmmm?
Turboprop. The only reason they didn't then was that they were still laboratory curiosities, and then mostly on the German side. BMW was working on them, and others. Fuel consumption argued against WWII-era deployment as well. They wouldn't have had the range. These days, assuming a prop-driven requirement for some reason, it would've been a turboprop.

However, the real answer to "What would an air-to-air/strike fighter look like if developed today?" is this:

#### Jay Kempf

##### Curmudgeon in Training (CIT)
I'm not sure there is a paying market for a new Merlin, yet. But the engineering design is already done and there is at present a lot of practical knowlege of these engines (and others); there are currently airframes pretty much being built from scratch and at some point someone will be willing to give making new engines a try (TVAL have already done this on an only slightly smaller scale). And if a detailed digital model existed then it would be a big obstacle out of the way. Casting pattern designs could be generated, tool paths for the milling, etc.
A digital model is so not the big obstacle! Metallurgy, casting technology, testing, tweaking, rebuilding cause all the castings cracked, reverse engineering, surface finishes, yadda... are the big obstacles. Anyone can make a geometry cartoon. Verifying the thermo alone is a huge deal. I am assuming there are drawings and hopefully specs to copy but getting a light weight (relatively) and highly stressed design like this right is all in the details. The digital model only gives you geometry and nominals. At the time the Merlin was built there were apprentice programs for machinists and bench fitters. That experience from what I can tell is LOST COMPLETELY. I deal with this on mere 30 year old cars and finding the original intent when reproducing one single part is not trivial and takes a ton of broad engineering and manufacturing experience that you can't learn from the internet or books. Aluminum engine blocks are fickle mistresses. You don't just dump aluminum in a hole, break the sand off and then machine it.

#### Jay Kempf

##### Curmudgeon in Training (CIT)
Turboprop. The only reason they didn't then was that they were still laboratory curiosities, and then mostly on the German side. BMW was working on them, and others. Fuel consumption argued against WWII-era deployment as well. They wouldn't have had the range. These days, assuming a prop-driven requirement for some reason, it would've been a turboprop.

However, the real answer to "What would an air-to-air/strike fighter look like if developed today?" is this:

View attachment 16750
That wasn't my question... My question was what someone would do if they had to design a P51 with a piston engine, with boost of course, today knowing what we know now. I know we would design a turbine for the same task today. But say you had to design a P51 but you had modern material science and all the collected knowledge on piston engines today. Would it be a boosted, reduction geared, lower cubic volume design? I think it would. Might even be a diesel.

#### Topaz

##### Super Moderator
Staff member
Log Member
That wasn't my question... My question was what someone would do if they had to design a P51 with a piston engine, with boost of course, today knowing what we know now. I know we would design a turbine for the same task today. But say you had to design a P51 but you had modern material science and all the collected knowledge on piston engines today. Would it be a boosted, reduction geared, lower cubic volume design? I think it would. Might even be a diesel.
Probably. You'd probably see higher compression, direct fuel injection, and so on, but really there isn't much reason to do the job differently in gross overview than the Merlin or Griffon. Obviously the electronics would be better. The airframe would likely be different. The Germans, and Dornier in particular, were showing the way forward for propeller-driven fighters when jets came on the scene and changed the equation forever. See some of the later projects, such as the Dornier P.247, P.252 and some of the later Jumo 222-powered Focke-Wulf projects. All were going to swept wings and pusher propellers.

#### Autodidact

##### Well-Known Member
A digital model is so not the big obstacle! Metallurgy, casting technology, testing, tweaking, rebuilding cause all the castings cracked, reverse engineering, surface finishes, yadda... are the big obstacles. Anyone can make a geometry cartoon. Verifying the thermo alone is a huge deal. I am assuming there are drawings and hopefully specs to copy but getting a light weight (relatively) and highly stressed design like this right is all in the details. The digital model only gives you geometry and nominals. At the time the Merlin was built there were apprentice programs for machinists and bench fitters. That experience from what I can tell is LOST COMPLETELY. I deal with this on mere 30 year old cars and finding the original intent when reproducing one single part is not trivial and takes a ton of broad engineering and manufacturing experience that you can't learn from the internet or books. Aluminum engine blocks are fickle mistresses. You don't just dump aluminum in a hole, break the sand off and then machine it.
You seem to have misunderstood what I was saying. The engine has already been designed. The specs still exist. A digital model will have to be made as the first step in manufacturing new ones economically. It's not THE big obstacle - it's A big obstacle. Getting it done would be a step toward eventual replication of these engines. It could encourage someone else, who has expertise in a needed area, to add their part into the process. I'm familiar with the various steps required in manufacturing alloy castings, including maintaining a clean melt, designing the runners and risers to avoid problems with inclusions, cold shuts, etc., heat treatment. And please don't accuse me of having said I can do all of those things, I am only pointing out that I also know that you can't just dump aluminum into a hole. It would be a journey of many steps, just as you said, but you can't go anywhere without taking the first step.

And it would not be a cartoon. And not just anyone could do it.

#### Jay Kempf

##### Curmudgeon in Training (CIT)
You seem to have misunderstood what I was saying. The engine has already been designed. The specs still exist. A digital model will have to be made as the first step in manufacturing new ones economically. It's not THE big obstacle - it's A big obstacle. Getting it done would be a step toward eventual replication of these engines. It could encourage someone else, who has expertise in a needed area, to add their part into the process. I'm familiar with the various steps required in manufacturing alloy castings, including maintaining a clean melt, designing the runners and risers to avoid problems with inclusions, cold shuts, etc., heat treatment. And please don't accuse me of having said I can do all of those things, I am only pointing out that I also know that you can't just dump aluminum into a hole. It would be a journey of many steps, just as you said, but you can't go anywhere without taking the first step.

And it would not be a cartoon. And not just anyone could do it.
Assuming the engineering is done and documented enough to actually reproduce one of these things then there is no reason to build a digital model. Any machinist worth his salt could just program right from the old pen and ink. But my guess is that the entire picture isn't there.

I could model an entire engine accurate to the nominals with all features probably in less than a month assuming I had the drawings to work from. At the end of that and with the drawings in hand I still couldn't make an engine out of it. So much time has passed that all of the details of how it was manufactured probably aren't there. Good experienced people can work backwards from implied intent but experimentation is still the only way to find out if it is going to actually work and make the numbers and reliability after the effort. Back then some 19 year old most likely woman was responsible for the melt and her secrets are still with her, and there was so much word of mouth stuff and training in those factories back then and it was all moving soooooo fast for the war effort.

But you are correct that it could be redeveloped. But it is a monumental task.

Sorry if I missed the intent. Sorry if you thought I was accusing you of saying that you could do all the steps. I actually can including the modeling, stress, I wouldn't do the thermo myself, I can do the casting and procedural design and I could do the pattern work. But it would take a team of 30 of me to actually do all that in a reasonable amount of time.

A digital model is a step. And I think people put too much weight in 3D pictures. I am sensitive to that because I am responsible for the 3D pictures. Show someone a 3D picture and they think the thing is done. Kinda goes like the old Homebuilt adage: you're 90% done, only 90% to go.

#### AVI

##### Well-Known Member
Thought I'd read somewhere that Jack Roush was either in the process of manufacturing, or had manufactured some Merlin components ....

#### Autodidact

##### Well-Known Member
Any machinist worth his salt could just program right from the old pen and ink.
True enough, that's how they were originally made (Edit: oops, sorry, misunderstood your post there). But don't you think it would be easier if you could use CNC to make patterns and possibly do some of the other things? A supercharger impeller? I haven't seen everything in a Merlin but I have seen some of it and there are some complex things there. Just making the casting patterns would be crazy hard, I think. Your drawing skills are great, but I think you're underestimating the time it would take a little bit, maybe. I didn't mean to come across as overly defensive, sorry. I'm trying to work on that.

Last edited:

#### Autodidact

##### Well-Known Member
Thought I'd read somewhere that Jack Roush was either in the process of manufacturing, or had manufactured some Merlin components ....
I read in one of the magazines that he was. Several years ago; an example was valve rockers (four valves, one cam) with tungsten wear pads. And welding up combustion chambers to install new seats and stuff like that.

#### Topaz

##### Super Moderator
Staff member
Log Member
I would think you'd need a full set of original drawings regardless of whether you intend to model them in CAD or not. Without all the original material callouts, hardnesses, surface finishes, tolerances, and so on, you're just going to build a fancy paperweight, not a working, reliable, engine.

Nothing against the modeling process - those paper drawings aren't going to last forever, so now's the time - but if you're going to do it, do it completely.

##### Well-Known Member
Could a new merlin be built from scratch today? Absolutely

Is it economical to do so? Not even close, nor will it be in the forseeable future

Here's a quote from V12 guru Mike Nixon regarding building new engines from about 10 years ago - adjust prices for inflation
"..had to have a propeller reduction gear cover, a casting about the size and shape of a bedpan, manufactured. “Pattern, casting, and machine work, it cost $20,000,” he says. “I look at that and laugh when people suggest building an entire new Merlin. It would cost$1 million per engine, easy.”

Aspiring cad modelers should be aware there are well over 10,000 individual parts in a merlin

#### Jay Kempf

##### Curmudgeon in Training (CIT)
True enough, that's how they were originally made (Edit: oops, sorry, misunderstood your post there). But don't you think it would be easier if you could use CNC to make patterns and possibly do some of the other things? A supercharger impeller? I haven't seen everything in a Merlin but I have seen some of it and there are some complex things there. Just making the casting patterns would be crazy hard, I think. Your drawing skills are great, but I think you're underestimating the time it would take a little bit, maybe. I didn't mean to come across as overly defensive, sorry. I'm trying to work on that.
I don't think I am underestimating the task. I actually have master pattern making training and have worked making patterns for both high volume and one off projects/companies. When I started with castings there were no computers so I have been there through the transition. Have done impellers both the old school way and with CNC patterns and hogged from billet. I understand the task. Drawing skills are such a minor part of the equation is the point I was trying to make. The experience of the factory that made these things originally is most likely lost in the majority. Sure there are probably some heat treat notes and tolerances on the drawings but that doesn't tell the whole story. These engines also came from a time before GD&T so there was a lot of holding your thumb up and interpreting fits. The guys on the machines worked from journeyman instructions. All that stuff is what makes an engine stay together or not. Been in production machine shops and foundries my whole adult life. Been around crusty old production managers and learned from old school engineers my whole career. Taught myself all the fancy new stuff and don't get me wrong I like the new CNC stuff. Most times when you sit down at the tube to design anything you had to have gone through all the compromises and done all the research to even start putting the first feature into a new part.

A parenthetic aside: when working on a german car you go get the factory manuals because they are available to us home wrenches by doctrine. Reading into the factory manuals you quickly realize that all the information you need isn't there. Why? Because the factory wrote those manuals for their factory trained technicians that have gone through their approved courses and apprenticeship program so they know all that stuff that isn't in the books. If you weren't part of that system you are lost unless you can find the journeyman that teaches the course or someone who has been there and done that by trial and error and/or research.

Having reversed engineered many, many, many products over the last decades I see this all the time. I get paid to figure all that stuff out. It is incredibly frustrating to think you have something complete only to then find some piece of critical information that negates all the decisions you made leading up to that point and having to start all over again. But that is how it is done.

Sorry if I have stirred up a hornets nest. I am very sensitive to people thinking that the pretty 3D model is done and therefore the project is complete. That is one of the things about the new computer way of engineering that I really don't like. People make inferences about the level of effort required to do something based on a single impression of a picture.

#### Autodidact

##### Well-Known Member
Sorry if I have stirred up a hornets nest.
No prob. I'm a little sensitive myself, lately. The thing I was thinking was that making the patterns by hand would be a BIG job and that being able to take surfaces off of the model and design patterns and cores and runners and what not to be made on CNC machinery would be a big help. Even if all that someone did was to create a set of new castings, which is not an engine by a long shot, it would automatically put some more engines into the air. It would also be a big step toward new engines. Actually, new castings was the main thing I had in mind as far as the modeling goes - the machined steel internals would benefit a little less from digitising. Those would in some cases be very difficult; the crankshaft was forged, wasn't it? But when an engine blows a hole in its crank case at Reno, it destroys the case but not everything. So there are good internals looking for a new home. New castings would be very helpful IF they could be done for less than a king's ransom.

If you weren't part of that system you are lost unless you can find the journeyman that teaches the course or someone who has been there and done that by trial and error and/or research.
There is a lot of knowlege about these engines out there since several firms are rebuilding and modifying them. The model isn't a panacea, but I just think it would be very helpful. Done correctly, FEA could be run on say, the crankshaft, and the suitability of other manufacturing methods and materials could be evaluated, such as a crank machined from billet instead of a forging. It could be very helpful in any case, I think. No substitute for the basic knowlege though, and I wasn't meaning to appear to suggest that. In an earlier life, I was a CAD draftsman - draughtsman even, I would occasionally do pen and ink stuff and most of my schooling was geared towards hand drafting. Personally, I much prefer hand made drawings, both to do and to look at. But I can see great possibilities for digital modeling as well.

Last edited:

#### Jay Kempf

##### Curmudgeon in Training (CIT)
There is a lot of knowlege about these engines out there since several firms are rebuilding and modifying them. The model isn't a panacea, but I just think it would be very helpful. Done correctly, FEA could be run on say, the crankshaft, and the suitability of other manufacturing methods and materials could be evaluated, such as a crank machined from billet instead of a forging. It could be very helpful in any case, I think. No substitute for the basic knowlege though, and I wasn't meaning to appear to suggest that. In an earlier life, I was a CAD draftsman - draughtsman even, I would occasionally do pen and ink stuff and most of my schooling was geared towards hand drafting. Personally, I much prefer hand made drawings, both to do and to look at. But I can see great possibilities for digital modeling as well.
I much preferred the quality of the presentation of pen and ink on silk. I think that one of the things lost on the new computer aided stuff is the beauty of line weights for dramatic affect. Thicker line weights for outlines of parts, thinner for dimensions and section lines were great. Getting the computer to do that now takes time we don't have. So we stuff things via email to the machine shop and get on the phone and clarify. Not like it used to be but we are in a much more competitive market than we used to be. I loved doing drawings on the board. But I love virtual prototyping. Things change. I build things. These are tools. The tools are better for moving things faster than they were when I started in the profession. But some things I miss. Some things amaze me. Emailing a file and getting a part back that is exactly like I wanted via Brown truck up a dirt road in Vermont ski country is a good development in the R&D world and constantly amazes me. It is a good time to be a technical person.

For the Merlin. A consortium of loosely affiliated technical people and motivated suppliers could do a lot of damage in a hurry. There are those like me that do this stuff every day and like being involved in cool projects that further the craft. The craft must survive at all costs whatever contribution is required by all of us.

#### dirk_D

##### Well-Known Member
Could you buy a jaguar v12 a cast reduction hub with similar styling?

#### craig saxon

##### Well-Known Member
I have a Merlin crank shaft which I saved from being scrapped some years ago. The same guy had rods, pistons and superchargers as well as a bunch of old airplane stuff including complete Armstrong Siddley Cheeter Radials still in their packing crates. When he died, his family just scrapped what they could and took the rest to the tip. Wish I had known when they were doing that. As it is I now have a rather large and heavy paper weight........

#### Autodidact

##### Well-Known Member
A consortium of loosely affiliated technical people and motivated suppliers could do a lot of damage in a hurry.

Or over a longer period, it could be one of THE open source projects of the century.

#### Dana

##### Super Moderator
Staff member
CAD modeling a Merline would be a big project, but it's far from impossible. I'm sure it will happen. Building it is another matter entirely. Could it be done, certainly, but at what cost? The production tooling was based on production runs of thousands of engines, with an entire industry geared up for that kind of technology (and, of course, the "cost no object" impetus of a world war). Today you might be able to machine a crankcase from billet, but the forged parts would cost you a pretty penny.

Today, of course, a turboprop would be the choice for a prop aircraft in that power range. If it had to be piston, the modern practice would certainly be smaller displacement, higher rpm, and higher ratio reduction gearing.

-Dana

Things should be made as simple as possible, but not any simpler.-- Albert Einstein