Four years ago I started modeling a 1989 Ferrari 460 Grand Prix Formula 1 car for fun. About a week later, for one reason or another, the personal project was abandoned. More than likely due to a growing workload at the ad agency I was working for at the time.
Fast forward to August 2016. While rearranging some files on my archives I stumbled across the files again and decided I would spend some time finishing it. It has been really interesting cracking this model open again after so many years.
Where I Left Off
Right off the bat I was confronted with a slew of nagging errors. It’s crazy how skills evolve subtly over the years. Proof that you learn so many things that you don’t realize in the moment.
So the new goal for this project is to not only complete the model – but also make sure the asset itself is as accurate as possible. I also intend to fix all of the errors from the previous attempt, like bad topology or the absence of natural part breaks.
I will be updating this post as I make progress. So here we go!
Starting Anew: Fixing the Starting Point
My first task will be to tackle the areas of the model that have bad topology or are incorrect to the real world car. A lot of the basic form had been laid out but a great deal of it needs to be redone.
In the past, I always strove to model with as few parts as possible. I’m not sure where I picked this habit up but I have since dropped it in favor of breaking shapes up. Also, my skills with defining subdivision shapes were still very new. As a result, the rims are a bit mushy in places and don’t have any of the fine detail modeled as separate objects.
Modeling pieces as separate objects is more realistic and gives you the opportunity to control where edge loops start and stop.
I am also going to spend some time adding more features like the axle, hubcaps, brake calipers and disks.
On the original model, there was a basic shape for the upright, but it wasn’t very accurate. Also, the wishbone and pushrods terminated as separate objects – roughly aligning with the port on the upright. More than likely due to the fact that I didn’t have a good solution for merging the geometry. These will all be remodeled and tightened up.
After a long morning, here are the final results for the wheels, uprights, and wishbones.
The Nose Cone
Next up is the nose cone. In the past, I ascribed to the edge loop “fencing” method. This combined with the desire to model all details on one solid object led to a lot of unnecessary edge loops spanning the entire length of the object. For instance, the nose tip has dozens of loops, all tracing back to detail being held in the side intakes.
The first step will be to break the nose cone off. This will allow for me to do part lines but also control the edge flow better.
The base model also has some bumps in the mesh when smoothed due to bad topology and will need to be resolved.
What you don’t know does hurt you! I remember being taught to keep detail as low as possible when modeling (pre-subdivision). I am guessing this is what lead to modeling the holes for the wishbones and push rods as 4 sided punches.
Four sided holes are very finicky and don’t resolve as easily to quads, so these will need to be switched to 8 sided holes.
Also … that edge flow needs to go.
In addition to the excessive edge loops contributing to pinching, there is a subtle pinch in the real world car that needs to be added to the tip.
Now with all these changes made, the nose cone is much easier to tackle. Here are the final results for the nose cone.
The Main Fuselage
This section turned out to be a big undertaking and took a little longer than I thought it would.
I started remodeling it as shells with seams and then decided to go back and redo it (and the nose cone) to be “watertight”.
There were two reasons for this. One, this would be realistic to the actual substrate, in this case, thin sheets of carbon fiber. Second, this would make for easier 3D printing if I decide to print it.
I first tackled the shape of the windshield and the surrounding cockpit. Things weren’t lining up correctly and they were not accurate to the real car.
The next task was to break the fuselage into its realistic sections, redo to topology and check for any inaccuracies. This took a bit of back and forth to finally get right. After scouring through my reference, I realized that the back half of the fuselage was completely wrong – which required rethinking the current design.
Looking at the reference I also realized that the intakes on the sides were actually going the wrong direction and need to be swapped. I took this opportunity to break the channels out as separate parts as well. This will help when developing shaders and will also open the for to doing cool exploded views of the car.
Finally, after about two days, the fuselage is in pretty good shape. There are areas that still need a little bit of work, but I really need a break from the fuselage! Here are some images of how it is coming together.
Coming Soon: The Aerodynamics
Let me know your thoughts in the comments. Is there something you liked about this article that you want to see more of? Prefer it presented differently? Have a cool technique you want to share? I’d love to hear your thoughts.