What is the correct way to scale a plane model, such as making it 100 times smaller? What should be done with the weight? What about Unity's precision limits for such things?
I have tried something like you recommended, but it is hard to get stable calculations. But after I use the water as a fluid (and changed density to 100) the x0.1 model of the glider started to fly like stable. Centre of mass should be corrected too.
Scaling an aircraft is quite complicated as different properties are related to the size of the aircraft in different ways. Mass for example is roughly proportional to length^3 through volume and density while most aerodynamics depend on area which is proportional to length^2.
Factors you will want to consider when scaling your model are the thrust to weight ratio and the wing loading of the newly scaled model. Keeping those ratios the same as the original model should give decent results. Bear in mind that moments of inertia will also change.
Single precision floating points are plenty accurate to work across a scale on the order of 100x.
Your main issue will be getting the wing loading correct with the new model. If you have a very small mass but very large aerodynamic loads then the simulation will become numerically unstable and won't be useable.
Hi Conor, thanks for your advices!
I have tried something like you recommended, but it is hard to get stable calculations. But after I use the water as a fluid (and changed density to 100) the x0.1 model of the glider started to fly like stable. Centre of mass should be corrected too.
Hi Eugene,
Scaling an aircraft is quite complicated as different properties are related to the size of the aircraft in different ways. Mass for example is roughly proportional to length^3 through volume and density while most aerodynamics depend on area which is proportional to length^2.
Factors you will want to consider when scaling your model are the thrust to weight ratio and the wing loading of the newly scaled model. Keeping those ratios the same as the original model should give decent results. Bear in mind that moments of inertia will also change.
Single precision floating points are plenty accurate to work across a scale on the order of 100x.
Your main issue will be getting the wing loading correct with the new model. If you have a very small mass but very large aerodynamic loads then the simulation will become numerically unstable and won't be useable.