Gilera Bi4 Nordwest 3d scanning/modelling

[pegaso_grigio]

Domani do un occhiata al Dakota...

Tomorrow I'll take a look at Dakota's carter

[Duranti]

Another update; this time the right hand side crankcase cover. 

This cover is by far the most complex part of the parts I've replicated in 3D, there are many features, curves, fillet-transitions, and surprises, it's very interesting to model something someone else has designed, I have learned alot from just this part!

The last picture is with the scanned part, and 3D cad part, intersecting, this is a very useful way to check deviation. Most high end foundries these days use 3d scanners to check the "as-cast" part against the theoretically correct 3D model.

Remember this is not reverse engineering, just an accurate (as possible) representation of the exterior of the Bi4, and either way it would be impossible to capture a 100% perfect design intent, only the original engineers know the secrets! 

Last but not least; wow, I am so impressed by the beauty (and of course the engineering) of the Bi4 parts! 

[Matte]

Very well detailed, superb work.

[Duranti]

Thank you Matte, I really appreciate your comments, this project is a big learning experience for me, and this cover especially was a challenge with all the details, drafts, fillets, but it turned out good!
At the moment I am locating all the small parts to complete the external surfaces of the engine, I am planning on scanning/modelling every part that makes up the outside of the Bi4, it's a big task, but as an enthusiast it is worth the effort. 
Next up will be the cylinder head, hopefully I'll have it ready in the next few days 

[yumax]

Hi Duranti, really nice job,
I'm worry to understand if it's really possible use the file you achieve for reproduce the elements by milling or CNC machine is it possible in your opinion?
Sorry for so stupid question but i'm really not expertise in this field.

Great! go ahed.

[Matte]

I don't think it's possible, he doesn't use reverse engineering. Practically, his software makes a 3D photo of the part but without getting direct measures of the item. Is it right?

[Duranti]

Good morning yumax and Matte 
Thank you for your messages!

Matte is correct, the objective of this project is not reverse engineering, the objective is to create an accurate representation of the external geometry of the Bi4.

However, 3d scanning is widely used for reverse engineering of cast parts. And in order to correctly reverse engineer a cast (and further machined) part, you would have to combine the 3d scanned information, with coordinate measuring machine (CMM) information.
This way you capture the "organic" information of the casting geometry (which can deviate from one casting to another) using 3D scanning, and "cnc machined" (same numerical info every time relative to a set datum point) information using CMM.
I only have a 3D scanner (with max 0.1-0.2 mm surface accuracy) at my disposal, so I am supplementing the 3D scanned information with critical manual measurements using vernier calipers/micrometers and so on.

The parts I'm modelling can be cnc machined, but the information of the machined areas that I've captured using the 3D scanner are not accurate enough to produce a perfectly fitting part. Also the geometry of the models have been optimized for the casting process, and it would not be a sensible starting point for a fully cnc machined part.

Almost anything is possible though with a 5 axis cnc milling machine, and super-finishing, but the cost would be a show stopper for small series part production. It would only be feasible if you are machining molds for a larger production run of castings. 

My application of the 3D models will be less critical, but I still need all the attachment points on the engine to be as accurate as possible, so I will be refining the dimensions of the 3D model at the end to get as close as possible to the fully assembled "real" engine. 

[Duranti]

An fantastic alternative to "machined from billet" these days, is to 3D print a positive of the part (the part itself), using a specialized wax type filament, to be enable creation of prototype/small-series investment casting! 

[Matte]

An fantastic alternative to "machined from billet" these days, is to 3D print a positive of the part (the part itself), using a specialized wax type filament, to be enable creation of prototype/small-series investment casting! 

This is a really nice idea.......

[yumax]

Hi Duranti, thnks a lot for answear and clear info.

Fantastic!!

[Duranti]

You are very welcome yumax 

[Duranti]

This time it's the top cover for the cylinder head, wow, not easy to replicate this part, the most difficult thing with reverse engineering is certainly how to identify geometry, and there are many intersecting shapes that terminate into eachother, I got some gray hairs from this one, haha    But it turned out pretty good!

Next up is the cylinder head itself, it will be an interesting (and demanding) part to model, stay tuned 

[yumax]

wonderful.

[CFASD]

Great works...
many years ago:
http://www.gilera-bi4.it/public/forum/index.php/topic,3596.msg36955.html#msg36955

but for time, work and budget, i stop it.
i buy David Laser Scanner, & Arduino for automatize the scanner, but i have no many time, so i stop it.

My dream of 3D scanner Bi4 models so i see in you, for many project in my mind, as for example, an 3D explosiv pieces pdf book, or others.

If you want condivide it, as it's the spirit of this forum, and that i substain, you can if you want, condivide your 3Dfiles in various format.

Thank f y attention.
'AO.

[Duranti]

Good evening CFASD,
Thank you for your positive message 

You have done fantastic work with your 3D models, as well as really nice rendering work, excellent! I am sorry your work stopped, but I understand that this type of work takes alot of time and dedication! Very interesting to hear that you used Arduino to automate the scanning process, I have thought of something similar, but never had the chance to try it.

I already invested around 200 hours (and many hundreds of Euro in software licenses) of work into scanning, measuring and modeling the Bi4 parts you see in this thread. It's very time consuming and specialized work.
Regarding the Bi4 3D models; I might consider sharing a model for public use, but I will have to see how sucessful to end-result will be.