F1 Exhaust System

[autogyro]

I am always a bit dubious when mathamatics and any form of simulation is used to prove how a technology in the real world actualy works Chris.
This is not a criticism of your posts at all.
I found the down side of taking the data gained as absolute, when Glasgow University did a computer simulation on how an autogyro flys and the CAA used their data to establish the current flying regulations for this aircraft type.
Nobody at the CAA has had any flight experience on autogyros and neither had Glasgow University. They ignored the previous data that we had gained from a full certificated autogyro development program under the old Ministry of Aviation from complete actual flight data recorded officialy using the 'Concorde' testing facility and other then official proceedures.
The present regulations on the CAA books for autogyro flight are totaly incorrect and dangerous IMO.
They have even agreed to let us undertake a new program to establish true regulations at our expense.
So you see why I am never completely happy with just math or simulation.

[xxChrisxx]

I am always a bit dubious when mathamatics and any form of simulation is used to prove how a technology in the real world actualy works Chris.

Well that's just because your main experiences come from a time when simulation was incredibly unsophisticated.

Simulation is really the only viable option in a modern world (increasingly so). It allows more options to be tested in a shorter time period. Things are never released without a validation test.

So with prototypes, you may be able to test 5 solutions from 40 fully in a given time period. With rapid simulation, you can test all 40, then put forward the 2 most promising ones for full development in the same time period.

A final design/test should always be a full practical test (just to make sure you haven't screwed anything up).



So whereas it might take one man a week to fully weld up and shape an F1 exhaust for testing. Pressure wave software and CFD will allow you to try many many configurations to get the best compromise between power and drivability, then send the best one to be made for testing.

I think the exhaust is now the only peice on an F1 car that is fully manufactured by hand.

[Just_a_fan]

I remember back when doing my first degree a lecturer asked a question (to do with smoke venting of a building as it happens). Everyone duly tapped away at calculators etc. and we all got the same figure. Pleased as punch until he said "no, that's the result of the calculation; what's the answer to question I asked?"

And that key distinction can be easily lost if one just reads the figures out of context - which is all too easy to do.

[xxChrisxx]

And that key distinction can be easily lost if one just reads the figures out of context - which is all too easy to do.

Especially with CFD. It's a bloody dangerous thing really, solutions can look close enough to be sensible (as they all look remarkable similar) but far enough out to cause disaster.

It's why nothing passes me personally without a full battery of hand calcs some thorough checking and doesn't leave the door without a proof test. Although I work with nice lumps of metal, so it's relatively easy to check.

[riff_raff]

xxChrisxx,

"Things are never released without a validation test."

You hit the nail on the head. Digital analysis and simulation tools (CFD, FEA, etc.) don't take the place of knowledge and expertise. They simply allow the engineer or designer to rapidly iterate many design solutions, and compare them on a qualitative basis, without the delay and expense of testing. These tools will make a competent engineer much more productive, but they cannot make a poor engineer more competent. And in any engineering discipline, whether racing or aerospace, analysis is always validated with testing.

"I think the exhaust is now the only peice on an F1 car that is fully manufactured by hand."

I would disagree. All composite parts (wings, tubs, engine covers, etc.) are still painstakingly laid-up by hand. And F1 tires are also constructed by hand. The tire construction and composite part layup processes still use manual labor because it is more cost effective and precise than automated techniques, given the low production rates F1 requires.

However, I'll have to admit there's nothing as beautiful as a TIG welded, 4-into-1, stepped diameter, Inconel, F1 exhaust header.

Regards,
riff_raff

[xpensive]

I am always a bit dubious when mathamatics and any form of simulation is used to prove how a technology in the real world actualy works Chris.
This is not a criticism of your posts at all.
...

You just insulted every living engineer, the ability to make up a theoretical model, plausible enough to be trusted is what makes an engineer, everything else, like validation-tests or what not, is just a technicians job-description.

I think.

[marcush.]

as always there is no black and white decision and it is always a question of where you are starting from and what you are trying to achieve.

going down a new road of development simulation and the lack of reference points will inevitably grow the cloud you are producing...the simulation is based on too many assumptions to be really exact.
If you are working in a field were all parameters are precisely defined and all mechanisms are undestood the simulation is just perfect to optimise.

Still neither testing nor simulation will produce ideas itself ..If you have creative thinking people with experience around ,the results and seemingly random effect wwill start them creating no ideas ,but in itself it is an evolutionary process at best.

with phsical testing you have of course the real big issue of repeatability ...and in the world of F1 my impression is those uncontrollable or not well known variables have a much bigger impact on the test result than the bits or iterations under test..at least at times..

[xxChrisxx]

You just insulted every living engineer, the ability to make up a theoretical model, plausible enough to be trusted is what makes an engineer, everything else, like validation-tests or what not, is just a technicians job-description.

I think.

Thats going a little bit far I think. It's not an insult to be weary of simulations, by themselves they can be very dangerous things (which is why they are always backed up with real tests). Engineers can go on to autopilot with simulation and overlook simple things.

Without technicians and machine shop workers the engineer can't realise his designs. I don't like engineers who look down on technicians as though they are doing an inferior job.

[autogyro]

+1 Chris
Some of the cleverest guys I have met in motorsport work on the shop floor.
Without them the engineers and designers would not achieve much.

[marcush.]

no ,the boffins are those who know it all,you must not forget this.
If it ain´t work ,then of course it is a unsolvable complexity and please don´t bore with basics ...
it does pay to know a lot ,but only if you are able to apply this knowledge to produce the desired results repeatably ....when the ifs and whens do avoid taking decisions you can as well throw it into the bin and take gut feeling decisions and rely on experience ..

[F1_eng]

I absolutely agree that technicians are very important and working with very skilled technicians is very pleasurable, particularly good machinists and fabricators.

I also agree what makes a good engineer is that fact that he can apply his/her theory to be useful in solving a task, not just know about a lot of things but can't actually use them. During my previous studies, when I was trawling through endless mathematics books for years, I was actually thinking and finding applications for all areas and new methods I was reading about. As an example, reading through books around vector calculus, I was thinking this stuff is brilliant so I decided to spend quite a lot of my time researching vectors and geometric calculus, for the reason it had endless applications and was so useful in modeling a lot of systems.
At the time, I wasn't particularly interested in deeply abstract mathematical concepts like, what is infinity, as in the real world, this kind of thinking is not the most appropriate. Now though, I find myself interested in these abstrat ideas, but only for stimulation of mind.

Mathematical modelling is completely amazing, I would say that the world would be a lot worse off without it. I fear that you guys haven't had the pleasure to have met and interacted with very good mathematical modellers, being able to start a piece of software is not what this is really about. It's about understanding how the models work, interact and the coupling between different routines, validation, FOSs.
There are things to bear in mind with any method, such as modeling turbulence is in most cases averaged, combustion modeling should strictly also be averaged because the behaviour is random to an extent. This is an example where by most users of commercial software don't fully understand the systems and the methods and if they do try and validate, the simulation is way off.

Just embrace the technology and when you don't understand something, ask....rather than get angry and start slating the methods.

[marcush.]

+1
I can only applaud for the stance you are taking with this.
in fact I´m absolutely sure it is pure gold being able to create a working and workable model of complex interactions .
My doubts start only were someone tries to sell a theory as a model of reality and the cloud he is drawing up is really bigger than what could be achieved with a bag of experience and common sense.
In fact I have build some exhausts based on calculations from simulations that really topped all my expectation by a long shot.

[xpensive]

What do you know, an eloquent engineer on F1T, thanks F1E, very neatly xpressed.

Having said that, my intention was never to knock the technicians on the floor, but practical know-how always stems from a theoretical model if you backtrack a bit. Xperience and conventional wisdom is useful, but would hardly produce much of an F1 car, would it?

[marcush.]

now you are insulting chapman and all the brilliant guys of the past,my friend.
F1 has become a boffins playing ground not long ago..

[xpensive]

Chapman was and is a myth, he never designed or engineered very much himself, but he had engineers like Marice Phillippe, Peter Wright, Tony Rudd and later Gerard Ducarouge around to do it for him.

Chapman himself was a conceptualist than anything else, who would have found himself completely lost in today's Formula One.