Delage 15-S-8 Grand Prix (1/8)

[Roy vd M.]

@CrazyCrank At least equally important to lubrication would be to use an abrasive, I think. But I will definitely keep White Spirit in mind, had not thought of that. Maybe a combination can be found in a cutting fluid (that I use with the lathe). 

 

Per PM I received the suggestion to use ceramic sticks (thanks for this and other suggestions!). That may be something to test if I can get a hold of those things (1,6mm. is smallest size, which is great, but I can only find them available in large quantities) and if I won't be able to find a solution for another problem: making the swirl marks on non-flat surfaces. That's going to be the greatest challenge, as I've been able to make 1,6mm. swirls on flat surfaces using sand paper. That works fine... but there are several non-flat surfaces. I will come back to that. 

 

The hereunder was meant as a 'part 1 of 2' post providing an overview of 'the swirl issue'. Part 2 is postponed now that I'm waiting for a possible solution to be sent to me... 

201. My intention is to replicate the engine d.d. 1927. So far I have not seen any 1927-dated photograph on which a 15-S-8 engine is depicted without the metal swirls. Here a 1927 photo:



202. In 1927 there were four Grand Prix cars of the 15-S-8 type (quite different from nowadays, where there are only two cars per team). I did not find any evidence for major differences between the four engine specimens. In 2017 there only one quite original specimen remains (chassis #1) but that car has obviously also been restored. Here a picture of the engine block of chassis #1 that received a beautiful coat of swirls during its restoration.



203. That's the only remaining engine block with this kind of swirls (except block #4 that was restored half like this, half 'clean'). The three other blocks look like this: 

 

Chassis #3 (the championship winning car of Robert Benoist), or what's left of it: 



204. Chassis #2: 



(Courtesy Wikipedia Commons)

205. Chassis #4: 



(File found here, copyright unknown, picture will be deleted upon first request of copyright holder)

206. After the 1927 Grand Prix season two more cars were built, ordered by rich clients of Delage. These cars are not as valuable as the 1927 cars for reference purposes (especially chassis #6 which was completely amended during 90 years) because they cannot show (origins of) the 1927 situation. The engine of #6 was replaced, but #5 looks like the other engine blocks: 



207. Interim conclusion: 

- I did not find any proof that one or more of the 1927 engines did not have swirls.

- Swirls applied in 1927 may have been removed through sanding and polishing. 

- Chassis #1 was restored that way (although the gearbox did not receive swirls, strangely). 

- Chassis #2 does not have swirls in 2017. 

- Chassis #3 does not have swirls in 2017 but that may be changed upon further restoring that car. 

- Chassis #4 only has swirls at the lower half of the block in 2017 (see photo above).

- Chassis #5 does not have swirls in 2017. 

- Chassis #6: block does not exist anymore. 

 

I do not know where they owners of chassis #2, #3, #4 and #5 got their info leading them not to apply swirls to their blocks (perhaps a conscious choice) and I can't verify this. 

 

Conclusion: until I found proof that in 1927 there was a Delage 15-S-8 featuring a swirl-less block, I'll keep trying to replicate them (until I'll go mad).  

 

208. How is the pattern applied in real life?

[video=youtube;NNb4AXDjnN8]https://www.youtube.com/watch?v=NNb4AXDjnN8[/video]

TO BE CONTINUED... WAITING FOR A TOOL TO ARRIVE, PERHAPS THE REMAINDER OF THIS POST IS NOT NECESSARY

[Codger]

Roy, please do not ban me from your thread for the following suggestion but a new thought has arisen.

 

In a word 'Decals'. I know, I know - nothing looks more real than real metal. But much like the carbon fiber decals of today, I have seen decals of several types effectively create the look of the 1:1 material. In particular, carbon, wood on instrument panels, and instrument faces themselves.

 

My humble suggestion would be to copy or photograph the actual pattern from a flat area of the engine. Scan and reduce that to scale size then print on decal paper. Admittedly, much simpler than spinning bits, lubricants and making thousand of swirls in ally if you DO find the proper bit and method.

 

Again, I understand your demand for perfection and natural materials but done correctly, I doubt you'd be able to tell. An added bonus is that it will work on (single not compound) curved surfaces.

 

-Your Humble Follower.

 

[PROPELLER]

Good luck Roy!
But you know my opinion about swirls... And for me this engine, a pure piece of art, without swirls is more pure. 
Dan.

[Codger]

Roy, I feared I had angered you    with my decal suggestion but I didn't wish to drive you away. Several of us have interest in a solution or choice, as Dan has suggested, on your part.

[Roy vd M.]

Not at all! I did mention the decal-option in the beginning of this topic and said it would be my last resort-option (if nothing else works). I'm sure I'll try a couple more stuff before I grab the decals. But I'm very pleased that you take the trouble of advising me. As is clear, i can use all the advice I can get; tried some more things and perhaps I found a door toward a solution. Very perhaps... 

 

Also to Dan, I agree that the engine without those swirls looks pretty impressive and it's a matter of taste. I have decided to depict the engine as it was in 1927 and unfortunately (?) I have not been able to find proof that there was any 15-S-8 engine without those swirls. 

 

Pretty busy once more at work and other stuff, hence no updates. 

 

PS Finally did some modeling today, by the way. I decided last Monday that, as my girlfriend and kid will be elsewhere this weekend, I'll have this Saturday to myself. I made this challenge to my self: build a model within 12 hours plus an obligatory 1 hour break. I started this morning at 9 and ended at 9PM, so I effectively used 11 hours (+ break) to get from sprues to model. It was a lot of fun and something away from all devising, researching, trying and detailing. Here the result:

 

 

 

Yep it's Airfix' fossile (1964) 1/76 Bren Gun Carrier & 6PDR Anti-Tank Gun. I had a splendid time, this was a great in-between thing to enjoy the hobby. 

 

Hopefully I'll be able to update on the 'swirl case' soon. Lately I've been trying 1,5mm. diameter rubber pads glued to a brass rod.

[PROPELLER]

I found this...
May be it can be useful!
 

http://autorestorations.co.nz/project/1927-delage-grand-prix/

 

And I suppose you have this (small) book. If not, I can scan it for you.

 

 

Dan.

[malpaso]

Roy, I'm sure you know this but the swirl pattern is called "engine-turning"; plenty of info and video about it on the web.

At 1/8 you may be able to just scale down the real methods?  In Harry Woodman's book on scratchbuilding 1/48 biplanes he suggests tiny bits of wet&dry and lots of patience!!

Good luck.

Cheers Will

[Roy vd M.]

2 hours ago, PROPELLER said:

http://autorestorations.co.nz/project/1927-delage-grand-prix/

 

This website, and many more, have been linked and their info used, in the opening post ('miscellaneous'). Thanks for the tip though. 

 

Quote

And I suppose you have this (small) book. If not, I can scan it for you.

 

Yes I have it. 

2 hours ago, malpaso said:

Roy, I'm sure you know this but the swirl pattern is called "engine-turning"

VonTrips told me before. That's what I have used as search term mostly.

 

Quote

At 1/8 you may be able to just scale down the real methods? 

Believe me that's what I have been trying. The problem is that each swirl should be 1,5mm. in diameter. That is very, very small. An impression has to be made on into the nickel. If it were a simple horizontal surface it would be doable but there are many curved and difficult-to-reach spots. So yes I need to scale down to the real method (=flexible pad) otherwise it won't be possible. But the tools don't exist and anything I made thus far has proven impossible or extremely difficult to work with. I don't give up easily though. 

[sharknose156]

1. Quick time model: laudable effort to stay in the game.

 

2. Engine turning; for the Mefisto dashboard ( yes am back at my bench ) i used a small pointed rubber tip on my dremel at 10 000 rpm for a second per "dot". Worked fine. It is the impression i wanted to create and not the actual turning at 1/12 scale.

it does not take too much time. the challenge is the regular lining. 

 

cheers

 

[Codger]

Dan you have provided magnificent pictures and I'm sure Roy has a vast reference collection. The view of the timing gears that drive the cams is to die for. Sheer mechanical excellence everywhere on this vehicle. Thank you for sharing with us.

 

Roy, I support and salute you for the foresight to select such a fabulous subject. I know you will make this a visual feast and capture the total beauty of the prototype.

 

I hope you can soon turn your full attention to bringing this to fruition. I will anxiously await its development.

[Buster99]

Hello Roy,

I have noticed your build and am quite impressed, very nice!

As you seem to look for a way to make swirls in aluminium I may offer some advice:

Swirls like these are often used in watchmaking and called "Perlage". Youtube has a lot of info on how this is done and watchmaking tools should help you in achieving your goal.

I have a couple of special holders for making "Perlage" in a small drill press, it is like a pin-vise that holds a cylindrical piece of polishing rubber, it should be available in different diameters and grits but is for creating small 'swirls'. You should be able to obtain these from places where they sell watchmaking tools.

Do a google search for "Perlage tools" and it should give you sufficient hits. "Boley", "Bergeon" and "Flume" are brands that sell the tools, as they are German/Swiss companies, look for "Stiftpolierer".

 

I hope this willl help you in your search,

Marc

Edited October 27, 2017 by Buster99
correction

[CrazyCrank]

Thanks a lot Buster99, Roy was looking for such micro abrasive rods, and I think he will find (and me too) the "precious" at Bergon's: 

www.bergeon.ch

They have abrasive rods of 1.5 mm diameter, in very different hardness.

I think they are convenient for this "perlage", but only on flat surfaces, not curved ones...

 

Edited October 28, 2017 by CrazyCrank

[Buster99]

Hello CrazyCrank,

 

Yes those are quite nice but they also have a different type with larger diameters that fit in a pin vise type holder. 

I have a couple of those and the softer types will work on curved surfaces when used with the right pressure and rpm.

My former supplier is no longer in business but the tools can still be found.

See https://www.boley.de/shop/3224.stiftpolierer for examples, it shows the rods and the holder.

 

It will take some practice to get good results on aluminium as it behaves different from gold and silver etc.

Also keep in mind that these professional tools can be quite a lot more expensive than one might expect, the benefits are that they are really good quality and will perform well and last long.

 

Best regards,

Marc

 

Edited October 27, 2017 by Buster99
addition of info

[sharknose156]

 

this is my attempt using a soft polishing tip.

For the car i am building, the Mefsito Fiat, the dashboard is so old and beaten that the 'turning' is fading and spaced, unlike a Bugatti bouchonage.  

 

IMHO, the aim at least for me, is to create an impression at the scale of 1/12 for an onlooker at about 20 to 30 cm ... good enough for me.

 

This is at a speed of 10 000 RPM placing the plate under the rubber tip for a very brief moment and moving toward oneself to see the line better. Easy do different sizes  as you can see above.

 

Cheers to all,

 

Roy, we look forward to your progress on this amazing project, even if at a slow pace, perfectly understandable.

[CrazyCrank]

3 hours ago, Buster99 said:

Hello CrazyCrank,

 

Yes those are quite nice but they also have a different type with larger diameters that fit in a pin vise type holder. 

I have a couple of those and the softer types will work on curved surfaces when used with the right pressure and rpm.

My former supplier is no longer in business but the tools can still be found.

See https://www.boley.de/shop/3224.stiftpolierer for examples, it shows the rods and the holder.

 

It will take some practice to get good results on aluminium as it behaves different from gold and silver etc.

Also keep in mind that these professional tools can be quite a lot more expensive than one might expect, the benefits are that they are really good quality and will perform well and last long.

 

Best regards,

Marc

 

 

Hi Marc

 

You're right about those last tools your link shows, but their diameter, 3 mm, is too large for Roy's purpose, I think...for mine anyway, unfortunately 

 

[Buster99]

Hi CrazyCrank,

 

The rods and holders should also be available in smaller diameters, I have a set of 2mm and 2.5mm myself. You will have to search for the right brand and sizes for your specific purpose.

I have seen firetrucks using patterns with really large circles (5-6cm) and I have seen minuscule ones around 1mm on watches.

 

@Sharknose156: Also a very workable solution, if it works for you it is great! Downside can be that changing pressure can result in different diameter circles.

The watchmaker tooling is intended for high quality watches and will ensure certain standards, for private purposes it all depends on personal criteria.

The sample you show looks quite good actually 

 

Best regards,

Marc

[Roy vd M.]

On 15/10/2017 at 6:52 PM, sharknose156 said:

i used a small pointed rubber tip on my dremel at 10 000 rpm for a second per "dot". Worked fine. It is the impression i wanted to create and not the actual turning at 1/12 scale.

it does not take too much time. the challenge is the regular lining. 

The problem I found with a rubber tip like that, at least for my not steady enough hand, is that the abrasion would not go as evenly as I'd like to. It also cannot be used on inward curvatures. The solution I found thanks to all you guys, really (but culminated into Buster99's fantastic advice) helps get very even swirl marks and would be a recommendation for a possible future project.

 

This all having said, I must admit that the swirls you have shown do look better than my previous tests. 

 

On 21/10/2017 at 12:15 AM, Codger said:

Dan you have provided magnificent pictures and I'm sure Roy has a vast reference collection. The view of the timing gears that drive the cams is to die for. Sheer mechanical excellence everywhere on this vehicle. Thank you for sharing with us.

Yes in the opening post these and all other pics that I found online have been indexed and shared. 

 

I agree that the view of the incredible gear constellation should bring a grin of excitement onto the face of even the least technically grounded housewife. It is just astounding how high the quality of this engine, with all its roller bearings (hence probably the to-die-for-engine sound) really is. 

 

Quote

Roy, I support and salute you for the foresight to select such a fabulous subject. I know you will make this a visual feast and capture the total beauty of the prototype.

I do hope my attempts will prove to be worthwhile and won't turn out to be a disilusionment.

Quote

I hope you can soon turn your full attention to bringing this to fruition. I will anxiously await its development.

Today, through a friend, I found a company in Amsterdam that sells these 'abrasive sticks' that Buster99 was talking about. And it works (see below). Also by matter of coincident, today I received a number of measurements from Florida. Not the engine ones yet, unfortunately. But we're getting closer to the point where I can start. Also unfortunately, some measurements do not seem to match those deemed to be correct by general opinion. I'll get back on that later. 

On 27/10/2017 at 1:31 AM, Buster99 said:

As you seem to look for a way to make swirls in aluminium

Not exactly: I'll be using large lumps of brass. The brass will be shaped and hollowed, to keep weight to a minimum (the model will be very heavy even then). Brass cilinders will be hard-soldered into the brass. The brass-silver solder combo will be electroplated (nickel), to get a uniform nickel colour. That will be used as a basis for engine turned swirls. 

 

Quote

Swirls like these are often used in watchmaking and called "Perlage". Youtube has a lot of info on how this is done and watchmaking tools should help you in achieving your goal.

I have a couple of special holders for making "Perlage" in a small drill press, it is like a pin-vise that holds a cylindrical piece of polishing rubber, it should be available in different diameters and grits but is for creating small 'swirls'. You should be able to obtain these from places where they sell watchmaking tools.

Do a google search for "Perlage tools" and it should give you sufficient hits. "Boley", "Bergeon" and "Flume" are brands that sell the tools, as they are German/Swiss companies, look for "Stiftpolierer". I hope this willl help you in your search,

You really helped me out there Buster99. After all those attempts I did... this is it! I was able to purchase some 2mm. and even 1mm. 'Polierstiften', around the corner here in Amsterdam. See here. Without your help this build would probably still be on hold or I'd still be looking for solutions. I simply cannot thank you enough. As a symbolic token of my gratitude I will engrave or write your name and those of others who have helped me in my search for this answer, into the engine inner wall. Nobody will be able to see it once the engine will be completed, but as the modeler's saying goes... we know it's there! Thanks again to everyone who participated in this quest!

On 27/10/2017 at 9:10 AM, CrazyCrank said:

I think they are convenient for this "perlage", but only on flat surfaces, not curved ones...

The ones I purchased today are flexible and can be used on curves, albeit very carefully. 

On 27/10/2017 at 9:28 PM, CrazyCrank said:

You're right about those last tools your link shows, but their diameter, 3 mm, is too large for Roy's purpose, I think...for mine anyway, unfortunately 

The shop I referred to is a paradise for (large-walleted) modelers. Many tools that I could instantly need and which I had no idea existed... Many punch & dies... it was astonishing. To give an idea on their assortment regarding these sticks: they are available in several diameters, levels of abrasiveness and flexibility... from 0,1mm. onwards to 0,2mm., 0,3mm... honestly great! 

 

209. Here an impression of how tiny the swirls are... at Retromobile someone deemed it necessary to touch the engine, which is not cool, but it makes for a nice comparison. 



210. Here goes... 10 months of thoughts, attempts at problem solving... is this it, finally? 



211. Yes, it really is! It even works great on the reasonably hard nickel.



212. It's about the utmost left two vertical rows of swirls:

 

213. The effect: shine and glitter as in the real car. I'm sure that if it works, this is going to be an engine block model with allure, whatever one thinks of the 'bling bling' (any opinion is justified there). It's almost beyond believe that this really is a 1927-original feature.  
 

 

214. These are the holy grails... a 2mm. abrasive rubber stick along with a 2mm. mandrel. 

 

 

[Codger]

They were surely insane in 1927 when they did this.

You will surely follow their path these 90 years later.

And congrats to Buster with the solution to the puzzle.

[Buster99]

Hi Roy, so glad it works for you and thank you kindly for providing the supplier! That will be very helpful. Your results look very nice and I will keep following your progress of this most interesting build. My hat of to you!

@Codger, I am just glad to having been able to contribute. It has been a long quest for me too, many years of trying and searching that led to finding these watchmakers tools.

 

As a sidenote, I was informed that the swirls were originally not even used for decorative purposes but as a means of providing mostly aluminium surfaces with a type of grooves that would hold oil in order to prevent rust and oxidation. Later on it became a more decorative method. (this apart from watchmaking were it has always been a purely decorative technique as I have also been told)

 

Best regards, Marc 

[Roy vd M.]

On 19/12/2017 at 11:52 PM, Codger said:

You will surely follow their path these 90 years later.

The path to craziness, yes!  

On 22/12/2017 at 10:53 AM, Buster99 said:

As a sidenote, I was informed that the swirls were originally not even used for decorative purposes but as a means of providing mostly aluminium surfaces with a type of grooves that would hold oil in order to prevent rust and oxidation. Later on it became a more decorative method. (this apart from watchmaking were it has always been a purely decorative technique as I have also been told)

Those were my thoughts too when I first posed my swirl questions almost a year ago here. One of the reasons why I insist applying them if I can. 

 

 

 

215. I consider it probable that Mr Jim Ison personally measured the car (chassis #1, currently in Florida, Revs Institute; Collier Collection) in 1964. Remarkably, two sheets with Delage 15-S-8 scale drawings can be found online. According to Daniel Cabart (his book 'Delage Champion du Monde' of 2017) the car was in the Briggs Cunningham collection from 1955 to 1986. Jim Ison was a more or less famous professional modeller at the time. I feel little trouble visualizing Mr Ison having been granted opportunity to measure the car integrally. I noticed that these drawings really match the real car's lining very well; as of now I have not spotted apparent mistakes. This is a picture of the 1/16th scale model made by Mr Ison; at that time the bonnets and radiator of the real car #1 weren't blue but bare sheet metal, in a recent restoration this was restored. 



216. It was therefore my hope that the measurements received from Rev's Institute would match the drawing. Alas, not in all aspects as it unfortunately now appears. 

 

Mr Ison's drawing are very elaborate, with several top-, side-, front- and rear views, views of the engine block and more. The last couple of days I had a labour-intensive struggle with these drawings. First the right hand side drawings were laid on top of one another, matching as well as I could through horizontal rotation, shearing, scaling et cetera. The individual subdrawings, laid on top of one another, can be faded in and out nicely. This can be seen in this 22 second video:  
 

 

217. After that I continued uniforming and 'overlaying' Mr Ison's drawings. Unfortunately not everything seemed in order, the track gauge of side drawing on the one hand and rear view on the other hand did not seem to match. See the horizontal blue guidelines connecting the tyres:  



So research- and amendment work had to be done. A good preparation is half the job. First I checked consistency of the drawings, measuring wheel height and width to see if those match. If not I'd rescale horizontal or vertical size. 

 

218. After hours and hours of amending, rotating, shearing, measuring, aligning and repositioning I now finally have a workable set of (overlayed) drawings. Indeed the horizontal and vertical size of the drawing were not accurate (approximately 1% discrepancy) Correcting that fortunately solved the problem described in paragraph 217.

 

I won't show the drawings in full resolution but if anyone has serious interest in scratchbuilding this car I'll welcome his PM, usually I wouldn't mind sending him the original Gimp file.  

 

Now I'll show the several layers in the end drawing. The first layer is made up by the several views of the body.  



219. Here the second layer: a number of cross sections of the body (the first three out of six drawings). Using a slid I can set how 'intense' a layer is to be displayed. In extremo, the first layer becomes completely invisible on the first three drawings. 



220. Third layer: the engine. The same applies here: I can set the level of transparency versus the other two (or one of those). 



221. For example here's the drawing of the front of the car, overlayed with the engine drawing. The blue guidelines were disabled. 



222. On top of that layer several cross sections of the engine can be displayed, among which this (rather deteriorated, but still useful) blueprint of the engine. 



223. Also the engine supports front and rear can be visualised. Here the front engine support: 



224. Scale has been set, in real life millimeters as well as in 1/8th scale. For example, see the measuring line (between the two crosshairs). According to the set scale the distance should be 91,2 millimeter on my model.



Through all this measurement-intense workout I found out that tyres on chassis #1 do not have the correct size. They are slightly too low and slightly too wide. That cannot be concluded by means of the drawing alone (compared to the measurements provided by Rev's Institute), but also by using photographs anno today vs. anno 1927. Compare the tyres on the two pictures below, the difference isn't great but it is there. I reckon in 1964 the correct tyres were still on the car.  





(Copyright Revs Institute resp. Prewarcar.com, educational and referential purposes only, pictures will be deleted upon first request)

[harveyb258]

Superb work Roy, as always! Are you going to have naked bonnets, too? They look wicked!

 

Cheers, H

 

[Roy vd M.]

@harveyb258 Nosiree! I intend to depict the car as it was in 1927 as much as I can, sorry. 

 

 

Regarding this post: most casual viewers of this topic are welcome to fast-forward to paragraph 227 to start. You won't miss much, a lot of text mainly  Also other large text parts in this post can be skipped, they mainly show progress in thoughts / research. To make reading easier I enlarged some texts that can be seen as summarising. 

 

225. The preparations have not been promising, at least until today (from paragraph 234). Small deviations are usually easy to handle for me, but here there was a bigger problem:

 

- The several engine drawings of Mr Ison do not fully match. The devations are not too intense, but... 

- I have just one (!) blueprint of the car, regarding the front of the engine. I'll come back to that blueprint later. Unfortunately it does not match Mr Ison's drawing... the difference is more than 2%. That's 2,5mm. on scale, such being unacceptable for me (meaning that I can't start the build until I'll have more certainty). 

- The blueprint seems to correspond with what I visually deduct from photos, engine-width-wise... but if the width would be correct, the engine would never fit into the car length-wise. Grr. 

 

To make matters more complicated, the measurements from the USA are totally different as well. I would check them later. 

 

Here the picture that seems to correspond with the one blueprint I have... which is not possible if Ison's drawing is correct. The yellow line represents the blueprint, the green line Mr Ison's drawing. 



It's theoretically possible the blueprint is incorrect... but what are the odds? Otherwise there are structural errors in Mr Ison's drawing. I couldn't imagine that (but wait.. more on that later). 

 

I sent an e-mail to the owner of chassis #3, in my neatest French, hoping to get the opportunity to measure the engine block. You never know... I'd drive to and from France with pleasure if that gave me certainty. If I won't get permission, I'll kindly request the curator of Brooklands Museum for permission to measure chassis #2's engine block. If both don't work out, there's always Revs Institute in Florida for electronic measurement.. the thing is, that's quite a journey for a bunch of measurements alone. 

 

226. I scanned a drawing from a 1949 book, the first (and only uncut!) edition of 'The Grand Prix Car' by Laurence Pomeroy. The book features drawings of L.C. Creswell. He drew the Delage in 1942, I don't know which chassis he based himself on; probably not #1 (which I'll also come back to later). 
 

Right away big differences with Mr Ison's drawing occur, such as the sloping bonnets and rounded shape used in Creswell's drawing. That's not correct. I found another one of his drawings of the Delage online; it differs somewhat with his other drawing. 

 

Mr Creswell's drawings proved to be incorrect in several aspects, compared to several reference pictures. 

All this until...

 

... until I found a drawing that I had marked 'not to be used' last January. Not to be used, because it did not match the blueprint. Unlike the blueprint, the 'forgotten drawing' depicts a bit of engine side view. 

 

I implemented that drawing as a layer in Gimp. Remarkably the heights of engine block bottom, engine top AND crank shaft appeared to match with Mr Ison's drawing. That gave some hope. But the compressor side view didn't match Mr Ison's drawing at all, so on the 'lost drawing' it was drawn way too large. Or...? 

I thought to myself, let's have a look at some reference pics of chassis #3. It appeared that Mr Ison's drawing really could not be right (see paragraph 227) regarding the compressor. What if Ison hadn't been able to measure / study that part of the engine? An indication is the number of bolts on the compressor's front face. I can now simply count them because chassis #3 shows them (after the major accident with #3 in the '30s). I also have pictures of chassis #1's compressor front face, made during restoration. But Ison wouldn't have been able to count them because he couldn't see them. He may have guessed there were 12, while in reality there were 16. Elsewhere Mr Ison's drawings are true as regards the numbers of bolts. 

 

That could be the reason for the discrepancy of that part of Mr Ison's drawing. But still... why would this 'lost' drawing contain better information than the real blueprint?

 

And then I realised all of a sudden that I had seen part of that blueprint Mr Pomeroy's 1949 book... and that the book contains the remark that 'the drawings' are made by Creswell. This most probably includes the drawing that I thought was a blueprint. Especially because it looks a lot like other similar drawings of front engine views of racing cars in that book. 

 

I should have listened better to Mr Daniel Cabart, expert on the 15-S-8 Delage and author of the book 'Delage Champion du Monde' who clearly said that no blueprints of this car exist anymore... anyway I now realise some very important info about these drawings. And if the other drawings of Mr Creswell were not correct, why would this 'previous blueprint' be? Could the 'forgotten drawing' be the accurate one, after all? Or would they both be faulty?

227. Plenty of text (also for my own memory), time for some images.

 

Notice that the compressor was depicted a lot smaller on Mr Ison's drawing, approximately just as large in diameter as the valve cover's raising. 



228. That isn't true to reality. See here the large difference in diameter (chassis #3): 



229. On the next picture the difference in diameter appears (of course) larger than it really is, but the picture unmistakably clarifies that the compressor almost touches a profile that's on the same height as the frame. That appears to comply with another drawing... we're coming nearer to the solution! Once more this is chassis #3: 



230. Temporary conclusion: Jim Ison drew the engine right grosso modo, except as regards the compressor; that part should be lower and larger. He probably didn't measure that because it simply cannot be measured well if the car is completely built up. 

 

What until today I saw as a blueprint or official drawing, I'll only partly use. Comparing this drawing to the other ones, the engine block looks like a sagged pudding: way too low. For the sake of clarity, the 'previous blueprint' is to the left, the 'lost drawing' is to the right. 

 

Here I aligned the underside of the engine, as well as the crank shaft, to those in Mr Ison's drawings as well as the 'lost' drawing. As can be seen, the 'previous blueprint' 's valve covers don't reach nearly high enough. I had this suspicion before, because of the really slim lining of the engine according to Mr Ison, but I wasn't able to find proof for it, even after comparing to several photos. Even if you have many photos, it's very difficult to get to conclusions on the basis of those photos alone (as long as you don't have measurements); safe clear differences on small parts or distances / surfaces, as mentioned regarding the compressor. 



231. This is what the compressor usually looks like... not easy to measure indeed: 

 

232. I had a rather rare bright moment. The side view-drawing of the engine shows compressor, gear assembly and one cilinder. What if I copied that cilinder-part seven times... my calculations proved correct when I got eight in total. Next I posed them together as well as I could. Even then the engine block would turn out to be even longer than Mr Ison's drawing would lead to believe. That agrees with my interpretation of the photos (but later turned out to be incorrect..).

 

I also found out that the compressor in Mr Ison's drawing was positioned a bit too far to the rear. Have a look at this photo:



Here I exactly (perhaps unintentionally) photographed both front wheels in one line without parallax. It appears that the compressor leans over the center of the front axle about 1/4. That definitely does not correspond with Mr Ison's drawing, he thought the front of the compressor was situated just over that axle's center. I therefore moved all of the engine a bit forward. The next animation shows this clearly. The cilinder-see-through (with low hanging radiator) is the 'forgotten drawing', the bolt-lined version is Mr Ison's. 

 

 

233. I forgot one detail though (thanks to Haddewade who reminded me that the 4th and 5th cilinder are a bit further apart than the other cilinders).

 

I put some more space between those cilinders. This implicates that the rear end of the engine block is too close to the firewall for sure. Let's make the engine a bit smaller then... trial and error etc. It now looks like this: 



On a side note: stitching frames of a video I made hovering over chassis #3 from aft to front, was unsuccessful.. alas.

234. Next I entered all the measurements I received from Revs Institute and I amended the drawing accordingly. It clearly shows that Mr Ison's drawing isn't as good as I had thought. And chances are that all sources in literature and drawings are completely astray. In my first post I summarized: 

 

Wheelbase: 2.500 mm. or 2.495 mm. /98,2 in. or 98,8 in. or 98,5 in. or 8 feet, 2 1/2 inches

Or, from small to large, 2.495,0 / 2.494,4 / 2.500,0 / 2.509,5 millimeters. 

The sources can be reviewed here. A total of nine different sources provide a total of four different measurements of wheelbase. Between shortest and longest there's a difference of 1,5 centimeters; ok no problem. But now Rev's Institute informed me that the -measured- wheelbase is 2.578,1 millimeters, or nearly seven centimeters longer than the longest wheelbase value mentioned in literature. 

Based on that info I started amending the drawing. The tyre diameter as provided by Rev's (73,66mm.) sounds a lot more like Blockley's specification of 75,6mm. (Blockley 5.00x19 tyres, see here) than Mr Ison's drawing would imply: 82,2mm. (more than 8,5cm. larger in diameter!! A huge difference of nearly 12%).

 

Having the current tyre size as a basis is more convenient than using the (probably) larger tyres that were fit in 1927, regardless whether Mr Ison's drawing is correct... quod non, because:

 

From there on the basic shapes of Mr Ison's drawing turned out not to be correct, such as rear, front, bonnet height... everything really. Looking in retrospect, Mr Ison probably didn't measure the car after all; he probably used photographs of the car combined with other, also erroneous, drawings instead. 
 

Beside this, it now occurred to me that chassis #1 is the only remaining car with tall rear, like the cars originally appeared to have; and in 1927 still had, probably... unfortunately it's difficult to say with certainty due to the unclear nature of the pictures. Probably several side view drawings were made with other chassis #'s as inspiration. 

Here the difference between Mr Ison's drawing and the drawing with newly acquired measurements... that difference is quite substantial, it now seems. 
 

 

235. For those who don't feel like watching the above video (black&white drawing: Mr Ison's version; blue-drawing: sketch according to Rev's measurements; the two tones of blue are meant to simulate a bit of depth):






Total build time: 47h. 
Total measurements study: 45h.

[NickD]

Roy,

 

7 cm sounds enormous. Assuming Revs know how to wield a tapemeasure, I wonder who first used a wheelbase of approx 2.5 m. The rest could possibly transcription. 7cm is not. Either whoever first settled on 2.5m did not know how to measure, or the method used was flawed (easy if taken from photos) or the car has been rebuilt differently at different times in its history. Your assertion that the drawing was derived from photos seems likely if the error is that large but is there any possibility that the car had a different size at different points in its history?

ATB

Nick

[Roy vd M.]

@NickD Ill update more in detail later. Some preliminary highlights:

 

- Re-measurement proved it’s 2500mm. after all. Other dimensions are okay.

- New drawing looks very much near the truth.

- Engine drawing redone, fits beautifully now.

- Have asked for permission to fly over to Florida to measure the car electronically for greater accuracy and more dimensions, as drawing of Mr Ison cannot be used. Yes I’m crazy. Will get an answer on this request in a few weeks.

 

More info, including very promising new drawing, to follow.

[Hannes]

Dear Roy , in case you don´t know yet : In the book : "The roaring twenties  , an album of early motor racing " , written by Cyril Posthumus you can find some interesting photos and stories regarding the Delages  , I recommend this book because it is very informative .  Many greetings !   Hannes