# Thread: Dimensions for non-linear geometry

1. ## Dimensions for non-linear geometry

Hey everyone. I am currently an engineering student with some experience in autodesk inventor and solidworks and I have a question for you guys my instructor just cannot seem to answer. When building parts with straight edges or linear curves I find the modeling and drawing methodology to be straightforward and simple. However, once you enter the realm of non-linear geometry this changes quickly and I cannot fathom how its done.

When engineers design and create car frames or fighter jet wings how do they dimension these models when there is so many organic shapes intermingling? I have played around with T-splines and made some organic models myself but when it comes to the drawing stage to put dimensions into the part I am at a loss as to how this is done. Does anyone have any experience in this field or perhaps an idea where there might be some resources on the issue? Textbook or otherwise? Thank you for your time everyone!

2. In the modern world where a 2D engineering drawing is required we create many section views along the length of the geometry. Think of a loaf of bread and each section view is where the slice is and that would be detailed. For this system to work there must be a common reference or zero location from which the dimensions and tolerances are established. This is called a datum. Additionally, there would be an engineering note on the drawing that would specify the dimensional conformance between the section views (slice point) indicating the dimension and tolerance conformance requirements.

Where a 2D engineering drawing is not required we use 3D model based definition methods. This is where the geometry is manufactured directly from the CAD model. We have technology that can read the 3D model and manufacture and dimensional tolerance verify.

3. That is an excellent explanation Kelly! Thank you! I now understand the process well at at conceptual level. I wonder though? Might you have an idea where I might find detailed tutorials, examples, or other resources of this process so I could start practicing them myself?

4. No, well not that I can publish on a public engineering forum.. Industry practices and process is the next level you will encounter on graduating. Conclusively, we are well educated in our engineering colleges and universities but there will be much to learn when you enter industry.

You will find industry is much more practical than college...

Originally Posted by x355
Might you have an idea where I might find detailed tutorials, examples, or other resources of this process so I could start practicing them myself?

5. As usual, Kelly has a spot-on answer. By way of addition, no matter what method or tools are used eventually it all gets down to a fixed datum and x,y,z, coordinates of discrete points. Those points are either determined by some mathematical equation, some rough equivalent thereof, or manually determined. Maybe its because I grew up in the home of an inventive design engineer, or maybe because I pursued the profession myself, but for whatever reason I have for a long time been very interested in the development of technological tools. One of the most fascinating was a MANUAL three dimensional coordinate measuring machine I saw in the engineering offices of the Auburn Cord Duesenberg Museum in Auburn, IN. It worked on the concept of the bread slices Kelly mentioned above. Picture a large U-shaped frame mounted on rails in the floor. It was large enough to fully straddle a car. Each leg of this frame included an array of parallel movable individual rods. First, the designers would model the new car body full size in clay. Once it was complete they would position this frame at one end of the car, called Station 1, extend each rod until it touched the clay surface of the car, measure its extension and write it down. Then they retract all the rods, move the frame along the rails to Station 2, and repeat the process, station by station. Eventually they ended with a large written database that fully documented the three dimensional surface of the car. This database then became the reference they used to fabricate the tooling to produce that shape. Similar processes were used for aircraft, submarines, etc. Often the "math" wasn't needed at all. They could just manually model the shape they need in wood, make a sand mold, and then cast it directly in iron or other metals. My compliments on a very good question.

6. Thanks for sharing you guys. This information has helped a great deal.

7. This is very informative. Thank you for sharing it guys.

8. I think one way to model 3D objects is revers engineering.Engineers take special photos from the specific object and then convert it to for example CATIA, then they use cloud of points and make the 3D model by some methods,then edited this and use it.Many softwares has 3D measurings too ,and tell the specific x,y,z coordinates.Also I agree with Kelly's answer too.