• Advertisement

Archived

This topic is now archived and is closed to further replies.

Physics Engine, How do I make one?

This topic is 6068 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

I want to create a physics engine for my scientific simulation. Its basically about systems of guided trajectories in coordinated non-trivial persuit of non-static air borne target. I ordered this book called Missile guidance and persuit: Missile Guidance, theory and kinematics. I wanted to know if there was a book dedicated to game dev. in this context, which is to say modelling the discrete domain. Or do we have tutorial here (the search feature seems to be lacking in many ways on this website). Anyway, I AM going to do it, but can anyone help me make it easy? Regards |IRR| -- The hungary Programmer - till yet.

Share this post


Link to post
Share on other sites
Advertisement
No game specific books or resources on this that I know of, although there are *tons* of resources on basic rigid body dynamics. See Chris Hecker's docs, for example, on www.d6.com. Essentially, when you simulate a missile you *are* simulating a rigid body. It is the forces applied----lift, drag, pitching moment, side force, rolling and yawing moment, thrust (and any moment associated with or caused by thrust), etc.----that make your problem different from just a general rigid body sim. Your forces are continuous---applied throughout the flight---whereas game rigid body simulations often have just impulses applied during collision. Also, your forces are highly dependent on the angle of the missile's orientation relative to the wind, and its rotational motion relative to the wind. Basically, it is the force calculation that makes the problem a bit more challenging than what Chris and other game developers typically talk about. You can simulate the missile to different degrees, including a virtual stability augmentation system (SAS).

The book you ordered may be useful---I've never looked at that one specifically. I can suggest some engineering resources that may be helpful, although not necessarily easy to follow:

"Dynamics of Flight: Stability and Control" - Bernard Etkin and Lloyd Duff Reid. A classic text. Focuses on airplane flight dynamics, but the governing equations are similar for missiles. The BIG difference for missiles is that the stability derivatives (read book to learn about these) are going to be VERY different for missiles and that's what makes the flight of a missile unique.

The American Institute for Aeronautics and Astronautics (www.aiaa.org) has a rather large number of technical and text books. They may very well have something applicable and current. (If my company's CRAPPY web connection would actually work reliably I'd look up some specific titles for you. But alas, the CRAPPY web connection does NOT work reliably and I cannot get to www.aiaa.org right now.)



Graham Rhodes
Senior Scientist
Applied Research Associates, Inc.

Edited by - grhodes_at_work on September 4, 2001 4:20:22 PM

Share this post


Link to post
Share on other sites
Can anyone tell me exactly what a "physics engine" is?

I''m guessing that people like to toss this word around, but that it likely just means: the code that figures out which computations to do and then does them.

Share this post


Link to post
Share on other sites
A physics engine is, generally, part of a game engine that provides a simulation/approximation of physics. Basically, a model of forces like gravity, friction, inertia and how they effect in-game models.

Most games have traditionally had very simple physics engines, but that''s been changing now that video cards are handling more of the graphics work, freeing the CPU for more complex physics, AI, etc.

Share this post


Link to post
Share on other sites
Dear Dr. Rhodes,

Looks like I got to know the right person! I have had about 3 courses in Differential Equations, so even though I am not a physicist, I know the mathematical tools that go into the design. Basically I think it involves 3 things: The mathematical modelling of the physical enviornment, the control aspects associated with real systems, and the discretization of the first 2.

I happen to have a great professor on campus who has had a PhD in missile fuel optimization -- intercontinental Ballastics. An engineering scientist specializing in integral analysis is also available. The discretization part is up to me. You can quote any resources you want, and the difficulty level will hopefully not be an issue. One piece of advice that I would like to have: Do you think its a good idea to do research in this military area, even though I have no plans to pursue it subsequently (it just offered a nice research topic), considering that I come from south Asia? Would this adversely effect my chances for persuing graduate level education at a place like TICAM UT Austin? Kindly advice.

Regards

|IRR| -- The hungary Programmer - till yet.

Share this post


Link to post
Share on other sites
quote:
Original post by Zawar
Dear Dr. Rhodes,



I''m flattered, but I am not a doctor! (I have a Master''s degree, and spent 5 years pursuing a Ph.D but did not complete it. My area of expertise was civilian airplane aerodynamics and design, with Ph.D work in computational fluid dynamics.)

quote:
Original post by Zawar
I happen to have a great professor on campus who has had a PhD in missile fuel optimization -- intercontinental Ballastics. An engineering scientist specializing in integral analysis is also available. The discretization part is up to me. You can quote any resources you want, and the difficulty level will hopefully not be an issue.



The only other resources I would be able to recommend is any NASA papers (search on the techreports page, http://techreports.larc.nasa.gov/cgi-bin/ntrs) on subjects such as forebody aerodynamics and missile flight dynamics. About a decade ago, there was quite a lot of research going on in the lateral control of flight vehicles at high angles of attack using elliptical forebodies, tangential blowing, forebody strakes/chines, etc. The work I''m aware of was focused on manned aircraft, though, and I''m not aware of current or recent work. This would obviously be on the aerodynamics side of things, not control or basic flight simulation.

quote:
Original post by Zawar
One piece of advice that I would like to have: Do you think its a good idea to do research in this military area, even though I have no plans to pursue it subsequently (it just offered a nice research topic), considering that I come from south Asia? Would this adversely effect my chances for persuing graduate level education at a place like TICAM UT Austin? Kindly advice.



I can''t really offer advice on this area specifically. I''m not sure what UT Austin is focusing on in their applied math department, but your chosen research topic would seem to be more appropriate for graduate research in an aerospace engineering department. North Carolina State University was one of the places where there was some forebody research going on in the late 80''s (www.ae.ncsu.edu), though they''re no longer doing much research in this area. The fact that you''re not a US citizen *could* affect your ability to work on certain related research projects, and civilian projects might be more appropriate if you intend to pursue your graduate degree here.

Graham Rhodes
Senior Scientist
Applied Research Associates, Inc.

Share this post


Link to post
Share on other sites
>>About a decade ago, there was quite a lot of research going on >>in the lateral control of flight vehicles at high angles of >>attack using elliptical forebodies, tangential blowing, >>forebody strakes/chines, etc. The work I''m aware of was >>focused on manned aircraft, though, and I''m not aware of >>current or recent work. This would obviously be on the >>aerodynamics side of things, not control or basic flight >>simulation.

Aerodynamics will determine the range of manouverability options that we have so I suppose this work is still relevant. Say, I am asking this even though this is not your field, but do know any place where I can get information on implementation issues related with realtime machine learning using Dynamic Programming. Because there was this paper at researchindex.org just on that. Sorry for calling you a Doctor, even though I am sure you aren''t anything less .

Share this post


Link to post
Share on other sites

  • Advertisement