Starting again

I got a yellow book from the UK.

No more guessing, no more excerpts from 1994 rules. This is the real deal.

One more time, templates based on the rules

This time I'm going to do reference images on planes

Make a surface

Apply material

I'll select the DS Star material

The material is way too small. So double click on DS Star Material. Change the Material Size value. Click on the Texture tab. Were going to change the image.

Point this to the image of the car. I edited this image to have inverted colors to be easier to see.

I did a little bit of formatting to the image to that this makes sense

Oh yeah, the image should be a perfect square to maintain proportions.

After aligning the image, I would say that worked out pretty well

Top looks as good as it's going to get. I've known for a while now that this image isn't perfectly symmetrical

Rear. The quality is what it is. This is a picture of a picture from a book.

Forgive me Catia for I have sinned

It's never the modeler's fault is it? It's always the software's fault.

Oops. I'm an amateur. I have a lot to learn about Catia v5 surface engineering.

NURBS surfaces are powerful and accurate. They do what you tell them to do. If you make crap curves you get crap surfaces. I watched a terrible Class A surfacing tutorial, where the guy created a wheel well with some atrocious gaps, but I did learn something. NURBS single span surfaces are most controllable.

I think that I've accomplished what a novice can do with the Generative Shape Design workbench. I want to stay in Catia, and mostly stick with feature based modeling, but I think I'm going to start over and primarily work in the ICEM shape design workbench, so I can use the tools found there to try to create Class A surfaces. This has been a long road, so the journey continues.

Change of approach

I'm going to design the nose - monocoque - cockpit - airbox in Autodesk Maya.

My logic is this:

I've been explicit modeling these parts in Catia. This was not so much of a conscious choice, as I didn't know all the possible methods of creating these parts. So I initially figured that the best solution would be to do everything natively in Catia. I've had two problems with explicit modeling.

1) It's hard
2) The original car was not explicit modeled in CAD software.

The fact that it's difficult to explicit model is not so much of a problem psychologically to me. Kudos to you Class A surface engineers of the world working in ICEM Surf or Autodesk Alias. I've found it difficult to learn, and conceptually challenging, but entirely possible with patience. My big issue at the moment with my modeling method is that I'm not moving forward maintaining the kind of surface quality I want. This car of course was the poorest performing F1 car of 1993. It started life as a clay model, was 3d scanned, and then maybe some of the surfaces were tweaked just a little bit. About roughly a billion times I've wanted to reach into my monitor with my hands and shape curves with my fingers. I suppose that's what learning to model is like?

So here's the big idea:

Box model the nose - monocoque - cockpit - airbox in Autodesk Maya (yeah, I know not an engineering software) Point is to recreate the clay modeling process that created the real car. I'll wind up with a polygon mesh. I intend to transform this into a single NURBS surface. I intend to save it as in IGES file. I intend to import this into Catia, and then chop it up as I need to. I intend to. We'll see what is possible with the time and software available.

Why Maya? Because I like this guy's tutorial. Pixelbahn. It's enough for me. From all the research I've done, there's not a gigantic difference in the work that can be created with 3ds Max, Maya, Lightwave, Cinema 4d. I saw something about it being a bit easier to make NURBS surfaces in Maya, but whatever, everyone has an opinion on the internet. Having a descent tutorial on an interesting subject is good enough for me.

So, I'm going to go back to the beginning with my scanned kinda blueprints in Corel and make some usable images out of them.

1994 Technical Regulations

Yeah, I know they aren't the '93 regs, but close enough for my purposes

I'm going to make my reference images this time in 1/10 scale

1. Width:
The overall width of the car including complete wheels shall not exceed
200cm, with the steering wheels in the straight ahead position.

5. Overhangs:
No part of the car shall be more than than 50cm behind the centre line
of the rear wheels or more than 120cm in front of the centre line of
the front wheels.
Furthermore, no part of the bodywork more than 20cm from the
longitudinal centre line of the car may be more than 90cm in front of
the front wheel centre line.

The centre line of any wheel shall be deemed to half way between two
straight edges, perpendicular to the surface on which the car is
standing, placed against opposite sides of the complete wheel at the
centre of the tyre tread.

6. Height:
Except for the rollover structures, no part of the car can be higher
than 100cm from the ground. However, any part of the rollover
structures more than 100cm from the ground must not be shaped to have a
significant aerodynamic influence on the performance of the car.
Furthermore, any part of the car behind the centre line of the rear
wheels must not be more than 95cm from the ground.

All height measurements will be taken with the car in normal racing
trim with the driver aboard seated normally.

So I'll make my drawing a little bit bigger than this

I'm going to show the grid

Wheelbase can be found on wikipedia. 3030mm

Side view of car regulations & wheelbase

With reference image scanned from book

Wheels. Front tires are Goodyear 25.5 x 9.5 x 13. Rears are Goodyear 26 x 13 x 13 (read off sidewalls)

In metric terms that is

Front: 647.7mm diameter x 241.3mm wide

Rear: 660.4mm diameter x 330.2mm wide

Something like that

Rake. Video.

Do I design the rake into the car from the beginning? I think I'm going to say no. I believe it'll be easier for me to keep everything from the reference plane (read floor) to the top of the wings at 90 degree angles. So lets modify the reference image so that the floor is parallel with the ground.

7. Aerodynamic influence:
Any specific part of the car influencing its aerodynamic performance:

- Must comply with the rules relating to its bodywork

- Must be rigidly secured to the entirely sprung part of the car
(rigidly secured means not having any degree of freedom).

- Must remain immobile in relation to the sprung part of the car.

Any device or construction that is designed to bridge the gap between
the sprung part of the car and the ground is prohibited under all
circumstances.

No part having an aerodynamic influence and no part of the bodywork may
under any circumstances be located below the geometrical plane
generated by the flat surface described in Article 3.3.

No part of the bodywork in front of the rear edge of the complete front
wheels and more than 25cm from the longitudinal centre line of the car
may be closer than 40mm to the geometrical plane referred to in Article
3.3.

- So the front wing endplates must be 40 mm above the reference plane

There are going to be some issues with this no matter how I do it. I don't have my data from Catia in this drawing. For example, I don't know what the bottom of the nose should look like. Anyways, this is all part of a learning process so whatever. I think this time instead of tracing the car in Corel and exporting the lines, I'm going to use the raw image. It's positioned well now, so I'll draw a big box around the car.

So I think now, I'll export this and open up the exported image in Photo Paint so I can crop it

The benefit of all this work is that I'll have two images, the side and the top, perfectly aligned for 3d modeling. Now on to the top view

What do we know about the top view.

1. Width:
The overall width of the car including complete wheels shall not exceed
200cm, with the steering wheels in the straight ahead position.

2. Width ahead of the front wheel centre line:
The bodywork ahead of the front wheel centre line is limited to a
maximum width of 140cm. Nevertheless, any part of the bodywork ahead of
the front wheel centre line exceeding and overall width of 110cm must
not extend above the height of the front wheel rims with the driver
aboard seated normally and irrespective of the fuel load.

3. Width and shape between the front and rear wheels:
The maximum width of the bodywork behind the centre line of the front
wheels and in front of the centre line of the rear wheels is 140cm.

So in front of the rear wheel centerline the max width of the bodywork is 1400mm

4. Width behind the rear wheel centre line:
Bodywork behind the centre line of the rear wheels must not exceed
100cm in width.

Unfortunately my top view scan isn't as good as the side view one, but it's close, oh well, what can you do

Cropped

To Catia or not Catia?

To use Catia to create the surfaces of my monocoque or not? That is the question.

I'm banging my head against the wall. I understand that I'm an amateur and I'm not a trained, professional Catia Surface Engineer, but seriously, I'm getting pretty repeatedly frustrated.

I think I did a nice job on this, but I still have clearly not class A surfaces

This situation is essentially repeating itself over and over again, and really slowing me down. I remember reading on a forum that your surfaces are only as good as the quality of your curves. My curves are clearly not good enough. I don't seem to have the skill at this point to make everything gel using the methods I'm using at the moment.

I don't doubt using Catia for parametric part design of stuff like the damper I did before, or the brakes, engine, the list goes on and on. Also, I'll need to have everything in Catia in order to convert my parts to use in CAM. (of course I don't have to, but I want to, this is my Catia learning project). Plus, assembly design in Catia? Brilliant.

But I jealously look at what people do in other programs, and provide excellent tutorials like this one:

http://www.3dm3.com/tutorials/lightwave/gp2_modeling/

or

http://www.cgarena.com/freestuff/tutorials/max/formula1/formula1.html

or this (wow!)

http://www.youtube.com/watch?v=MGtNt3e2hCk&feature=related

So, the question becomes, if my method is really difficult to learn, and not yielding results, how about changing it? For lack of a more scientific word, I want to make the airbox pretty. Some of the parts of this car require to the millimeter precision, like the tires and breaks. Some of them don't, they just have to look right. It's hard to make them look right the way that I'm doing things in Catia right now.

What program to use? I've seen some awesome, truly stunning models in Autodesk Maya, Autodesk 3ds and Lightwave. But these are all polygon models. I'll have a hard time importing those surfaces into Catia and physically make something out of them. From what I understand, I want NURBS surfaces, and it's going to take some hardcore software to create them and then export them in a format that I can import into Catia without losing a ton of quality. First choice would seem to be ICEM Surf. Dassault Systemes bought them and integrated some of the functionality into Catia. But I think I might be going in circles using ICEM, because I'm already having problems with this in Catia. So the other logical choice is Autodesk Alias. We'll see. At the moment, I'm leaning towards trying to learn box modeling in Maya since there are such detailed tutorials on youtube.