Fiat 806: research and scratchbuilds

[CrazyCrank]

Dear Olivier

 

I'm not sure at all you were right, thinking so !

IMHO, the real car's photo isn't at the same scale than your model: 0.5 or 1% less

Edited February 5, 2017 by CrazyCrank

[Totally Mad Olivier]

Ok CC I don't pretend it is 100% right. I consider  an error margin of 1% as acceptable. But I go on thinking this comparison is useful for me and I hope it will be for you too...

Edited February 5, 2017 by Olivier de St Raph

[Roy vd M.]

Olivier please be very careful when comparing photos like that. Lens distortion in the 1927 photo will for 99,9% sure be very different from the lens distortion in your own photo. The photo was taken at the same angle (=same point of view), which makes for a good principal comparison, but the distance between camera and car wasn't the same, the scale isn't the same, the lens wasn't the same. 

 

That means the results are unreliable. 

 

You say an error margin of 1% would be acceptable, but in my estimate the error margin you're looking at is 5% or more. 

 

 

[CrazyCrank]

6 hours ago, Roy vd M. said:

Olivier please be very careful when comparing photos like that. Lens distortion in the 1927 photo will for 99,9% sure be very different from the lens distortion in your own photo. The photo was taken at the same angle (=same point of view), which makes for a good principal comparison, but the distance between camera and car wasn't the same, the scale isn't the same, the lens wasn't the same. 

 

That means the results are unreliable. 

 

You say an error margin of 1% would be acceptable, but in my estimate the error margin you're looking at is 5% or more. 

 

 

 

That is exactly what I had planned to tell him this morning, and, stupidly, I didn't dare to. The photographer is placed some meters (between 3 and five, I think), before the car

 

Edited February 5, 2017 by CrazyCrank

[Totally Mad Olivier]

Ok guys, I hear you don't agree and of course I fully respect your opinions. Thierry, about the fact the camera was before the car, obviously, I tried to have the same view angle.

So for you it is just a coincidence if all lines match and if one doesn't, giving me precisely the same impression I had when I compared my build with photo 1. For me, it is not a coincidence.

Using Roy's word, let's agree to disagree on this.

P.S: I can go up to 5% margin error (but not more )

Edited February 5, 2017 by Olivier de St Raph

[Roy vd M.]

Let's also agree to disagree that all lines except one match. I see three matching and three non-matching lines.

 

I don't want to discourage you, hopefully you'll trust me on that. My warning is not meant to annoy you or set you back. 

[CrazyCrank]

Just take a look, please on the picture heerunder

I've taken Olivier's picture and modified it with Photoshop

The lower part of his picture (The real car) has been resized to 101.5% from the original, and then placed under the upper part (his model)

The front and rear pixels of the resized part have been lost while saving the picture, because Toshop save only what fit on the defined layout

 

 

What can we see ?

- the center blue lines match

- The first blue lines before the center line nearly match

- the second lines before the center lines don't match, BUT, if you  draw a line joining the front axle of the real car and the first hole in front of frame rails of the model, this line will be vertical: so, they match

 

IF these considerations would be sufficient to affirm that the scale difference  between model and real car is exact, then:

- the model's cockpit is frankly too long

- the model's tail is too short (very few)

- the model's rear louvres are too backward

 

We are going to become mad with this car !

 

 

 

 

 

Edited February 5, 2017 by CrazyCrank

[Totally Mad Olivier]

17 hours ago, NickD said:

I was reluctant to "like" the post, because you were doing something so distructive

Dear Nick,

before taking such a decision, I tried (unsuccessfully) to get the good shape (with the good distance between the fairing and the body. Indeed, the necessary decrease of the fairing in height (see above) had made indispensable this modif (red arrows). 

Be sure I would have avoided that new "surgery" with new time lost if I had the choice. 

Notice that on the photo 3, we can't see the jonction left side. But the logic is that the fairing was set on the body (with rivets) as Hannes suggested with right above in the thread, and the best way to reproduce that area was to make a continuous body and set the fairing on it, as it was on the real car.

 

Edited February 6, 2017 by Olivier de St Raph
correction

[Totally Mad Olivier]

21 minutes ago, Roy vd M. said:

I don't want to discourage you, hopefully you'll trust me on that. My warning is not meant to annoy you or set you back.

Dear Roy,

I have absolutely no doubt about that! 

In fact, I would prefer you (and Thierry) to be right, it would mean in my mind I don't need to make a new correction on the rear part of the cockpit...

[Roy vd M.]

Coincidentally, as Thierry, I did some checking by software. Not using Photoshop but Gimp. 

 

Ironically two lines that I thought did not match, eventually matched perfectly because Oliviers vertical lines were not perfectly vertical. If made vertical, they match! 

 

 

(The second '0%' marker takes the position of the louvres as a basis of measurement)

 

The conclusion based on the above Gimp research is that the front side of the car is a bit too much too the front, the end of the cockpit is too much too the back and the tail of the car is a bit too much to the back too. 

 

But as said before (I refer to the picture with the face comparisons) if I were you I would take those 'conclusions' lightly because different lenses and lens distortions makes the comparison impossible. That would only be different if we know exactly what lens Photo 4 was taken with, if we were to correct that through software, and if we were also to correct the lens distortion in your own photo. Further correcting scale and distance, we would be able to make an accurate comparison. But unfortunately that's impossible. 

 

On a more positive note I think you totally nailed it! The shape of the car looks spot-on to me and I am sure (almost) everyone agrees. At one point that is arguably the most important factor: does it look like the real car? The answer is, yes it absolutely does. Try to get back a few steps, to Protar's kit. To use a famous millionaire and philosopher of human rights... Protar kit is a total disaster. Total disaster. Your amendments made the kit great again. True. 

Edited February 5, 2017 by Roy vd M.
Oops, picture without guidelines...

[Hannes]

Photo 8 attracts my attention more and more . It seems , two photos were used  ( look at the left rear wheel !) and these photos did not have the same scale and angle.

After this " collage " was done , an untalented "artist " tried to " improve " this composition by adding wrong details .

But we can use this photo partially , if we know , what´s wrong and what´s right !  Many greetings !  Hannes

[Roy vd M.]

6 minutes ago, Hannes said:

Photo 8 attracts my attention more and more . It seems , two photos were used  ( look at the left rear wheel !) and these photos did not have the same scale and angle.

After this " collage " was done , an untalented "artist " tried to " improve " this composition by adding wrong details .

But we can use this photo partially , if we know , what´s wrong and what´s right !  Many greetings !  Hannes

 

Hmm it seems that I forgot to list Photo 8B in the opening thread, from which it becomes all the more clear that Photo 8 is indeed edited.

 

Does anyone remember where that other version of Photo 8 was put? It's somewhere in the thread... with exhaust stains covering the rear '15' and taller louvres. If found I'll add it to the opening post.  

Edited February 5, 2017 by Roy vd M.

[CrazyCrank]

9 minutes ago, Roy vd M. said:

 

On a more positive note I think you totally nailed it! The shape of the car looks spot-on to me and I am sure (almost) everyone agrees. At one point that is arguably the most important factor: does it look like the real car? The answer is, yes it absolutely does. Try to get back a few steps, to Protar's kit. To use a famous millionaire and philosopher of human rights... Protar kit is a total disaster. Total disaster. Your amendments made the kit great again. True. 

 

I totally agree with that, @Roy vd M.

 

There comes a time where one does stop with margin corrections, to join @Hannes poetical/sculptor point of view: a right shape, a magnificent piece of art, either for modeller's eyes than for lay's ones

 

[sharknose156]

Dear Olivier  

 

i think you are spot on. I don t think there is any significant distortion if one is to use eyeball engineering and proportions ( as Hannes does (?)

 

1. Bordino looks perfectly human and normally proportioned and not distorted (perhaps less than 1 pct but does not look 5 pct as in the pic of Roy !? )

 so the middle part of the cockpit should be ok as a reference. May be not the front and back in same pct but the wheels are circular ... so it should be less than 1 pct 

 

Your car is in front of your naked eye and it is therefore real unless one goes into Kant and Kafka.

So it is not distorted.

 

So IMHO congratulate yourself !

 

The only small difference is the lenghth of the 3 louvres which is an easy fix IF you wish

 

The position of the louvres higher and very slightly further back is not significant !! IMHO

 

So no need for further surgery doctor 

Just finish the fairing and let s see the wheels on !@

[Totally Mad Olivier]

Thank you Roy, Thierry, Harvey and Sam for your compliments and encouragements.

Indeed, I will not modify the position of the louvres and I have ever increased a bit their lenght, on CC suggestion.

I have changed my mind and chosen 0,5 mm thick alu foil instead of plastic, for the body shape under the fairing. The alu keeps its form when you bend it much better than plastic:

 

[Hannes]

Another issue I found out is the height of the car over the street. On drawing 2 the whole car is deeper imho compared with our car .I believe , another tyre sizes or a delayed axes-height was provided for the successor car . It also would make sense , because the weight -reduction of 40 kg seems have to been achieved by removing parts at the bottom of the car . ( struts under the engine , the whole bottom panel,triangular support parts for the last cross-beam ) This way the reduction from 0,89 to 0,84 square meters of the frontal surface  in combination with an altered fairing could have been calculated as Rogliatti writes in his article .  Hannes

[sharknose156]

About the weight reduction on the next version of the 806 ;

 

there is a relation between weight and rigidity in racing cars, sometimes the compromise does not mean the most rigid car, so removing the triangle supports of the cross beams may have been to reduce weight at the expense of frame rigidity.

 

Also the type of circuit is important.

 

 If i remember the next races would have been Spain and England Brookland.

 

Brookland was an oval and required a very special tuning, may be slightly less rigid than on a twisted track, for the long straights and long curves and bankings, all turning in the same direction ( except at the fork ) and weight distribution more on the 'inside' of the Brookland steep bankings, like US oval racing, to keep the car from understeering and refusing to turn (like a swiss cow galloping).

 

these guys were for real.... not like most racers today complaining about safety and wanting automatic start devices...

[Hannes]

Dear Sam , there were also different dampeners for the leaf springs imho ! If you compare those as shown on drawing 2 with the photos , they are altered in dimension and shape  !

Maybe also the leaf springs were replaced depending on what circuit came  next ! It´s like formula 1 today ...

Liebe Grüsse !  Hannes

[sharknose156]

6 minutes ago, Hannes said:

there were also different dampeners for the leaf springs imho

 

 

Yes absolutely my thinking as well.  I think you are right.

 

The car was prepared for another type of circuit than Monza where it rained (soft suspensions required) . Whereas Spain would have probably been sunny and required more rigid springs. (leaf)

Edited February 5, 2017 by sharknose156

[NickD]

Olivier,

 

I'd just like to add my voice to the others. I am not sure why you prefer the pictures to the lines and drawings we prepared earlier. The pictures are very difficult to interpret accurately. The lines we have generated over the last few months are as close as we have got. They won't be right everywhere but they are much more consistent than the photos. If we use the lines and drawings, I can correct the errors (such as Hannes' comment above about the dampers). If we don't use them we will just go round in circles.

 

My view, some time ago was that the lines you had were really very convincing.

 

Regards

 

Nick

 

[Hannes]

I agree with Nick . Foundations are the frame on drawing 2 and some important mounting points .The main dimensions and shapes of the body can be seen on Nick´s wire-frames.

The photos are the final instance for fine-tuning. I started with the photos because I´m used to do this by profession. I also know about the traps if you have to judge a form by a 2-.dimensional picture.

Straight lines seem to be curved and vice versa. And even if I made many things right , I still must correct till I get my desired result.

Look at the tail´s end on photos 1 , 2 and 3 for instance ! It´s lenght cannot be judged correctly , because it bends around the corner and there are foresighting effects too !

If you look at photos 4 and 8 it seems to have a different lenght : In my Protar - scale the lenght of the tail´s end seems to be correct ! ( but not the situation of the rear axis of course ! )

It´s all relative and if you get closer to a photo just 1 cm , the whole scale and so the dimensions will change ! And the human eye can get  betrayed easily !

Many greetings !  Hannes

[NickD]

Olivier,

 

I drew up an interpretation of the shape of the rear spring carriers. The bolts are notional. They are unclear on Drawing 2. They are long which suggests they may have had some sort of lubrication attachment. That is what I drew. The nuts probably had split pins but I was not clear on the exact geometry so I left them out.

 

 

 

 

I used Drawing 2 and Photo 2 to estimate the shape. There is a positional mismatch - equivalent to 12mm full scale (FS) longitudinally. Haven't checked why. Could be several things.Drawing 2 seems to be simplified for the swan-neck.

 

The swan-neck shape appears to be tapered circular.

 

It is impressive how slender this is considering how loaded it is.

Hope it is useful

 

Nick

[Totally Mad Olivier]

Thank you very much Nick for this shape of the rear spring carriers, that will be very useful for me very soon (one of my next step).

9 hours ago, NickD said:

I am not sure why you prefer the pictures to the lines and drawings we prepared earlier. The pictures are very difficult to interpret accurately. The lines we have generated over the last few months are as close as we have got. They won't be right everywhere but they are much more consistent than the photos

I have been using both drawings and photos for my build. It is true that these last times, I used much more photos. I will soon make a comparison with Drawing 2, and if you want to help me for that, I will be very grateful to you, Nick.

 

All the best

 

Olivier

[Totally Mad Olivier]

15 hours ago, Roy vd M. said:

 

(The second '0%' marker takes the position of the louvres as a basis of measurement)

 

The conclusion based on the above Gimp research is that the front side of the car is a bit too much too the front, the end of the cockpit is too much too the back and the tail of the car is a bit too much to the back too. 

Dear Roy,

I don't understand the values on that photo: is it 2,25 mm, 3,37 mm and 0,75 mm? I suppose so. Thank you anyway for that study. I will soon make a check with Drawing 2 amended. I suppose my frame will indeed be a bit too long (probably around 3 to 5 mm). It will be much more difficult to have a certainty about the body behind the cockpit area with Drawing 2...

[Roy vd M.]

5 hours ago, Olivier de St Raph said:

I don't understand the values on that photo

 

The values are based on wheelbase of 2.400mm. and, given lens distortion, approximate. 

 

From left to right:

 

• -22,5mm.: on the model, the crossbeam section (a bit to the right of its center) would be 22,5mm. too far to the front. In 1/12th scale that is 1,9mm.
• -22,5mm.: on the model, the beginning of the cowl would be 22,5mm. too far to the front. In 1/12th scale that is 1,9mm.
• 0,0mm.: the top of the fairing matches exactly.
• +33,7mm.: on the model, the end of the cockpit (begin of 'rear fairing') would be 33,7mm. too far to the rear. In 1/12th scale that is 2,8mm.
• 0,0mm.: the position of the louvres matches exactly (note: this doesn't mean the position is de facto correct, especially taken into account the +33,7mm. mentioned above)
• +7,5mm.: on the model, the end of the car would be 7,5mm. too far to the rear. In 1/12th scale that is 0,6mm. 

 

Another comparison is more visual and provides more information: not only length but also height. Markings 1-6 make clear where the differences are. The 'video' first slowly slides from your photo into Photo 4. After that it switches between them six times, for easy comparison between the six markings. 

 

 

 

For example section 6 would be 94 millimeters lower on the model (true scale).That's 8 millimeters in scale 1/12. If the car was 96 centimeters tall (which according to amended Drawing 2 it approximately was, so give&take...), this difference would be 9,8%! 

 

But, again, on a more positive note: don't worry too much about these differences because what you see is not reliable.