My CAD models

[notjustshrawan]

awesome effort...try modeling a Ferrari..you might end up famous like this guy and his version of a Ferrari...
Ferrari F460 Tifosi Concept - Noeman GSM
check it out..

[Knuto]

Havnt done much lately.

I just started working on an engine analyzer. I want to model pressure in the engine piping and calculate thermodynamic and gas properties in the entire engine. Ive read the book Design and Simulation of Four Stroke Engines by Gordon P. Blair, which I now call the bible. Amazon.com: Design and Simulation of Four Stroke Engines [R-186]: Gordon P. Blair: Books

I have a problem solving one problem though. I have 4 nonlinear equations I need to solve for 4 unknowns. I get good result through Newtons method if make a good first guess, but it quickly goes bananas. Any mathematics around who can help??

Ive started modelling outflow from a constant volume cylinder. I attached a plot from the result so far. It shows the pressure ratio (p/p0) for cylinder/plenum as a function of time. Also shows leftward wave (going towards the plenum) and rightward wave (going towards the pipe end) and the superposition pressure of these at a certain point along the pipe. It is wrong in assuming that the pressure at the throat of the pipe is equal to the cylinder, so the cylinder emptys to fast. I need to solve the 4 equations to get the pressure ratio at the throat. But the good thing is that pressure waves dynamics in the pipe is correct.

Cyco, modelling your V12 performance will be my big aim. But its gonna take some time...

EDIT: After working a little bit i think the problem is the equations. I think one of them might be wrong. the method seems ok

[Alastor]

Originally Posted by Knuto

I have a problem solving one problem though. I have 4 nonlinear equations I need to solve for 4 unknowns. I get good result through Newtons method if make a good first guess, but it quickly goes bananas. Any mathematics around who can help??

EDIT: After working a little bit i think the problem is the equations. I think one of them might be wrong. the method seems ok

If you are still having problems consider giving Excel Solver a try, assuming an iterative solver will work for you application.

[Cyco]

Kunto, I can't help on the maths because while I think I understand what you are trying to model I'm not sure of what to expect.

Have you tried the Lotus freeware engine modelling tool?

[Knuto]

Originally Posted by Cyco

Kunto, I can't help on the maths because while I think I understand what you are trying to model I'm not sure of what to expect.

Have you tried the Lotus freeware engine modelling tool?

Thanks for the link. I think I have tried it out before
I'm trying to get a deeper understanding of what goes on within the engine. So far I've learned a lot. I get a big understanding of what effect each parameter has on performance.

Decided to relax at home this evening. I worked a little more in Matlab on the engine analyzer. Fixed a lot of small problems. For the equations I asked about, I just put a couple of minuses here and there and now it works.

I tried to model an example from the book, similar to what I posted earlier. Its emptying of a 912cm^3 cylinder of 1.5 bar, by opening a sliding-plate valve for 0.008 s. The pipe is 5.9 m long. Attached are to plots. First shows pressure 0.3 m from the cylinder and the second shows pressure 3.3 m from the cylinder.

The book says that 86.6 % of mass is still in the cyl after the valve shuts. I got 0.5% more than that, but I'm getting closer.

EDIT: OK, after reviewing the maps of discharge coefficients wich I copied from the book, Im now within 0.039%. I think the code is good now. I will go on to modell a 2-valve single cylinder next

[jugga]

its probably very expensive but i have used this engine modelling software called ricardo wave at uni. its pretty good

[drakkie]

Originally Posted by jugga

its probably very expensive but i have used this engine modelling software called ricardo wave at uni. its pretty good

The calculations done above, I have done as well if I understand it good. I had to do them on paper though and turn them in for school.

I couldnt find anything on it but perhaps a cheap/free student version is around of these programmes ?

[Knuto]

Originally Posted by jugga

its probably very expensive but i have used this engine modelling software called ricardo wave at uni. its pretty good

The calculations done above, I have done as well if I understand it good. I had to do them on paper though and turn them in for school.

I couldnt find anything on it but perhaps a cheap/free student version is around of these programmes ?

Thanks
This is what car-manufacturers use, right? Is it 3-dimensional CFD? Looks pretty advanced. Ive seen some CFD software with moving mesh capability, but it was really slow. What engine did you model?

Drakkie, what exactly did you calculate? Maybe you can help if I get stuck, because I havn't taken any classes in this subject. Do you have many classes about engines?

[Knuto]

Was able to model a 2-valve single cylinder today. The engine is just pumping air right now, will take care of combustion next time. Attached a plot of the results. Left is pressure in exhaust port, and the right shows pressure in inlet port as a function of crank angle. Its not very fast right now (59 seconds for 2 cycles) , but I will try to simplify it when everything is working. Maybe it will get a little faster then. For the moment I have 188 different variables and matrices spred over 900 lines of codes, so it starts to get tricky juggling them around.

[Klemmel]

When will I be able to test-drive a Knutomobil?? Seems to me your knowledge have covered every important field in automotive engineering, design, engine and chassis.

Keep up the good work and at least let me ride shotgun in it

[Knuto]

When will I be able to test-drive a Knutomobil?? Seems to me your knowledge have covered every important field in automotive engineering, design, engine and chassis.

Keep up the good work and at least let me ride shotgun in it

Hehe. When we're a 100 years old!
I was working a little bit on the computer tonight and thought I'd share some pictures.

Here comes a little history of the car on these pages. It started life as a radio controlled car when I was about 12 years old (first pic). In ~2000 the game Gran Turismo 2 was the big hit and I was very inspired by the Suzuki Escudo. My dreams evolved into a light, 4wd, quad-turbo V4 engined, minicar (pic 2-4 and 6). As time passed, the car got some more curvature and the turbo fascination faded (pic 5).

[Knuto]

The 50s Grand prix cars inspired me next and the car became rounder and less dramatic. But I went back and forth between front (pic 1-2) and rear (pic 3-4) engine position. I stopped drawing at this point and it didnt change until I started drawing in Solidworks. Meanwhile my big love for engines grew and I drew quite a few engines. I attached a pic of combined block/head.(which I think I posted a pic of from SW).

I will update when I get to draw some more, either SW or on paper

EDIT: sorry for poor quality

[Knuto]

A long time since I posted in this thread. I had almost forgotton about it.

I thought I'd post some pictures of the car I've been working on for about six weeks. I have a summer job in a really quit town so this was a perfect opportunity to finish the car.

Its kinda like a Targa Florio racer with a screaming V12. The goal is to model it in the game GT Legends (pic 5 shows the first working model). I have a deadline to finish it so many of the recent parts (diff, gearbox, valvetrain) are low quality models

The car:
Length: 380cm
Height: 96cm
Width: 164cm
Weight: 600kg (I think I will manage a couple og kg below 600)

Engine
60 degree V12
Capacity: 1998 cm3
Max power: 275hp at 10500rpm
Max torque: 195Nm at 9200rpm

The engine model consists of ca. 550 parts and the rest of the car is about 250 parts

I'm hoping to finish the cad modeling quite soon and then go on to use the models in GTLegends

[RacingManiac]

fan of the pullrod layout I see...

are the suspension kinematic generated through something? The front seems alright, thought the rear looks kinda funny, the camber curve looks to be iffy if you were to layout the inboard pickup that way, though can't say for sure without seeing the outboard location.....also not a fan of machined billet a-arms.....tubes are much easier to make and can be just as light if not lighter...

[Knuto]

I love the pullrod layout. I've simulated the suspension geometry. The camber gain at the rear is about 18% lower than the front, so not too much(?). It was a good compromise between rollcentre and camber gain. I tried different layouts but this combo felt the best. I'm no expert though

The wishbones were rushed As I said before I'm on a deadline with this car. The model don't show it, but they're supposed to be forged, not machined. I agree, don't like machined.

The car is not production-friendly If you get what I mean. If I were supposed to make this car I would be making many changes to make components simpler to fabricate