Mercedes or BMW Inline-6 Info

Well, maybe not quite ‘your own design’ yet, but the step just before that? LOL I have been gathering pictures & documentation on WW1 era inline 6 engines along the lines of the Mercedes DIII or BMW III. From what I gather (maybe wrong) they were quite similar, BMW superseded Mercedes inline-6 lineage as higher cylinder arrangements were then developed. I have a soft spot for these engines & the planes they powered. Either engine would certainly make a nice model built in quarter scale or larger. Some construction aspects would be more challenging compared to my Ohrndorf-5 radial. Like that long, spindly crankshaft, the bevel gear driveshafts to overhead cam shaft & the water cooling details. But we don’t improve unless we try, right?

I have not seen many (any?) running models of these engines for whatever reason. If you have come across any examples, please pass on the links. I suspect the typical FS welded/cast cylinder/head assemblies add complexity & don’t lend themselves easily to miniaturization. OTOH, similar issues occur on typical water cooled auto-style engines & there are many fine examples of those running in the wild. So, I thought I’d start this post to collect any info that might be helpful. I’m familiar with 3D CAD & more recently acquired a 3D printer which has proved to be quite valuable for pre-machining prototyping. That would make a great start I am prepared to embark on.

I’ll start Q&A discussion with the water cooling details because that is the biggest mystery to me right now. Maybe someone has access to better drawings or a dusty book on their shelf? Both Mercedes & BMW cylinder liners had an annular space for circulating cooling water around them. That could be accomplished on a model with an external jacket of some sort, maybe pressed in place or brazed on. There is also a water tube segment welded to the lower cylinder jacket on RHS. The ends extend fore/aft a bit so they can be connected together in series using rubber sleeves or similar. This tube could be replicated by brazing on the water jacket.

But cooling water also flowed through the heads in the cavity space between the valves & plugs & combustion dome shell. Both engines had aligned water ports on the front & rear of each head, presumably to be interconnected in series. The head's water cavity becomes trickier to miniaturize without jewelry grade casting ability or 3DP metal heads (hold that thought for now). Perhaps a modelling simplification could be: make a conventional ‘solid’ head with the normal features & drill a fore/aft water passage hole between the valves & attach extending pipe ends? This would not achieve the same wetted area as flooding an internal cast cavity, so less cooling effect, but it would be ‘something’. At least mitigates welding or casting.

What I am struggling with is how cooling water actually circulated from the pump, through liner jackets & head area & radiator. I can trace some pipe paths in some pictures but not the whole picture. Obviously some of the hoses have been removed on display stand engines but even the drawings are not giving me the complete picture. Perhaps circulation mode varied among engines? Give these a pictures/markups looksee & I welcome your comments.

Water pump is positioned at lower rear of engine driven off CS. I think it shows discharge into rear-most head in & the heads are interconnected via ‘upper’ water line (my light blue shade). For distinction I shaded the lower water line dark blue, which inter connects the liner jackets. It appears to be dead-ended at front of front cylinder? Presumably radiator at front of engine?

So, what is the water flow path? Cool pump water into all 6 heads in series? Does that gravity feed through each liner & exits via the lower jacket tube? Wouldn’t water be getting hotter & hotter as it progresses forward from cylinder to cylinder? maybe missing something obvious but not seeing what I would expect as return piping on left side.

BMW left side

Mercedes - looks like the pump is discharging into the lower tube line from the right side? This made me wonder if water was flooding up each cylinder jacket into to the head & cascading forward through the head tube line?

Mercedes? from a YouTube animation. Not sure about technical accuracy although some details look quite plausible. Suggests the water pipe of head closest to prop connected to radiator

(Refabricated) Mercedes from Vintage Aviation showing their take on cylinder/head

Wikipedia Mercedes

That is an interesting engine design.
The cylinder walls are very thin, no doubt to minimize engine weight.

Do we know if the full sized engine cylinders were cast ?
It looks a bit like the water jacket is a slip-on affair over the cylinder and head (perhaps this was already stated).

This would be quite difficult to cast to try and match how thin the cylinders and heads are.
Additional material could be added to the inside of the cylinder head, with smaller valves, to allow casting.
Ditto with the cylinder; reduce the bore and make the wall thicker.

How is the cylinder head fastened to the cylinder ?

.

CI, Thus far I have not focused too much on the FS construction details other than skimming some of the documentation I've gathered. There may also have been differences to how Mercedes went about it vs BMW. It is of interest to me too, but mostly for knowledge sake. It would be near impossible to reproduce it in a functioning scale model.

But coincidentally I did recall reading something in a BMW extract (for you) & may have stumbled on an answer (for me) regarding water cooling. As highlighted, they say the pump has 2 outlets for upper & lower. On another picture I just saw one (the main?) one which added to my confusion. Now Mercedes may have had a different approach and/or specific to their own inline-6 models.

I downloaded this book & translated to English. Surprisingly it said little about water cooling but I can look for other details. They were quite tight lipped about many details.
https://bmw-grouparchiv.de/irc/resultlist/detailpage?id=3052421

Excerpts from the BMW book. Much of the content was maintenance procedures but I appreciated the nice drawings. Just wish they had a top view. Some of the valve rocker box mechanism was only apparent after absorbing Mercedes information.

The Wikipedia link was brief but interesting. I had no idea a BMW IV inline-6 variant flew to 9,760 m (32,000 ft). I had no idea pilots were huffing oxygen to that extent in 1919.

https://en.wikipedia.org/wiki/BMW_IV

Some crude translations

As a model engine will spend most of it's running life at tick over with the occasional blip of the throttle cooling becomes a lot less of an issue than the real tghing that was running at higher revs under load for extended periods. So a reduction in wetted area in not going to be such a worry.

Other model engines like the 1909 make do with an anular ring bored around the cylinder liner and various drilled and sometimes plugged passage sto form the linking water passages.

As for the heads look at models like the Centaur & Wyvern or the Titan 50 where a pocket is machined or cast into the head and then closed in to form a water space within the head. I suspect the head cooling could be done without the pipes doing anything and some drillings from the cylinder jackets up into a void in the head would be more than enough cooling for a ticking over model.

To me the cylinders lend themselves to a silver soldered steel fabrication then thickness and all the external details are not an issue, cast iron liner to give a good bore and close off the water space. Casting would only be worth it if investment cast and you still nee dto make the pattern and waxes. 3D printing in stainless or aluminium would also be a viable option again with the iron liner.

Don't discount modern adhesives such as loctite to retain cylinder liners rather than pres or shrink fits or the likes of JB Weld to bond in the tops of cylinder heads, more than upto the temperatures and there will only be a few psi of back pressure.

Really comes down to wanting to make a working model in a reasonable time or wanting to spend an inordinate amount of time on details that won't be seen and will unlikely affect the running of the finished engine which to most will look exactly the same externally.

Edit having read the bit about the top & bottom inlet and chlinder construction then maybe go for all steel turning the "forged" cylinder with some external bands and then silver solder on a thin "tube" to replicate the welded on steel liner. You then have the option to stick with a steel bore like many a model aero engine or add an additional iron sleeve

Hello Peter,

As you know, I am a long time student of Mercedes Benz engine design and manufacture. MB were very conservative, preferring to stick with a construction technique, which was known to work, rather than innovate. This was particularly true with the welded fabrication technique they applied to their cylinders. The cylinder welded fabrication was used on the individual cylinders of the Mercedes DIII in 1914 through the 4 cylinder mono-blocks of the Silver Arrows race cars of the 1930's through to the early Mercedes W196 of 1955. (That's over 40 years). Thereafter they finally changed to the 'normal' water cooled combined crankcase and cylinder block with detachable cylinder heads.

Looking at the single cylinder welded fabrication by Vintage Aviation. I believe their 'take', as you call it, is entirely faithful to the original engine construction. The cast cylinder head is screwed onto the cylinder barrel before the two thin steel waterjacket halves are hand welded in place. Effectively making it a one-piece, non detachable, cylinder/ head. Each a work of the welders art as there were so many protruding parts to seal. Crosthwaite and Gardner used an almost identical technique to reproduce the 4 cylinder monoblocks for their replica W125 engines. You can see more of their process in Post 220 of my W165 build log.

The water plumbing is rather ambiguous in those historic photos you have found. There are certainly detail differences between the BMW and Mercedes in that area. My take is the coolant flows from the water pump outlet to the interconnected pipework low down at cylinder jacket level. The coolant then flows upwards (in parallel) through the six individual cylinders blocks, where it exits and collects in the interconnected pipework running fore and aft on the very top of the cylinder heads. At the rear of the engine, the top pipe connects to the input of the water pump and recirculates flow to the bottom of the cylinders. It also connects to the water radiator at the front. Where the outlet of the radiator connects back in is not clear.

Achieving the welded water jacket would be difficult at model scale. you would maybe have more of a chance at 1/3 scale than at 1/4 scale. As Jason says, you could possible get away with water cooling only the cylinders and avoid the complication of the cylinder head water jacket. Alternatively you could consider the electroforming the complete water jacket in copper as Stephen Wessel did on his 1/2 scale ENV V8 engine.


These are the naked cylinders


And these are the cylinders with the electroformed copper water jacket in place.

There is not that much info on the net, regarding how Stephen set about creating the copper water jackets. You would need to talk to him directly. Stephen and I are long time friends, I could ask him to see if he is willing to talk about the technique with you.

Cheers 

Mike

As far as a crankshaft for a running 6 cylinder engine, you might want to look and George's 300 6 cylinder engine, or any of the Shillings drawings.  There are many running Shillings out there. 

Dave