Latest revision |
Your text |
Line 12: |
Line 12: |
| Gain Commit access ASAP<br> | | Gain Commit access ASAP<br> |
| Work out how density is handled during simulation and add it to the mass/volume calculations in a more accurate fashion. | | Work out how density is handled during simulation and add it to the mass/volume calculations in a more accurate fashion. |
− | WOrking out how to best calculate the coefficient of drag for objects.
| |
| | | |
| Coding Environments Status:<br> | | Coding Environments Status:<br> |
| VS2010x64 : Success, no bullet.<br> | | VS2010x64 : Success, no bullet.<br> |
− | Fedora 17 KDE-Spin : Success, Bullet running. | + | Fedora 17 : Success, Bullet running. |
| | | |
| | | |
Line 67: |
Line 66: |
| 18th - 22nd : Continued work on gravitational function. Working out how to intergrate the changing angular and linear velocities into the bullet simulation, since the brlcad system would ideally recalculate this every step. | | 18th - 22nd : Continued work on gravitational function. Working out how to intergrate the changing angular and linear velocities into the bullet simulation, since the brlcad system would ideally recalculate this every step. |
| | | |
− | 23rd - 26th : Finalising the gravity calculations and working on the fluid system. Need to work out suitable intervals for these calculations. 60 times per second would be a massive performance hit, even with the calculations themselves being relatively simple. | + | 23rd - 26th :Finalising the gravity calculations and working on the fluid system. Need to work out suitable intervals for these calculations. 60 times per second would be a massive performance hit, even with the calculations themselves being relatively simple. |
| | | |
− | 27th - 29th : Finishing off the gravity, videos coming soon (with highly exaggerated values to make it obvious, movement is negligible with small mass objects, or large distances).
| |
− |
| |
− | 30th : Continue working on fluid. Namely how to best teach the system to use the best coefficient of drag for a shape.
| |
− |
| |
− | ====July 2012====
| |
− |
| |
− | 1st - 2nd : Integrating into system with a similar method to gravity, alter the acceleration for the rigid body then use that to change the velocity every step.
| |
− |
| |
− | 3rd - 5th : Learning about how to go about creating video footage. Will need to save a render of every step of the simulation, I think. Also worked on testing the drag functions.
| |
− |
| |
− | 6th - 7th : Changed how mass is handled in the simulation, added new parameter to rigid_body structure to handle this. Temp using bb_vol assuming density is 1 to calculate mass, will be changed to something more precise soon. Also working on making sure the drag function gets the right area, and started to try and calculate any rotational forces around a rigid body
| |
− |
| |
− | 8th : Tested basic drag and gravity functions with rest of simulation and values output are correct with my calculations. Made progress with video. Still working on how to calculate the surface area for drag function, as well as how to get a reasonable coefficient of drag.
| |
− |
| |
− | 9th : Accidentally broke my fedora VM by running a simulation with 1 million steps as an experiment. Files were backed up, currently setting up another one, trying out the KDE distribution of fedora. Issue with building that I will fix tomorrow (fb.h, unknown types window, XVisualInfo, and a couple of others on line 136) and then I will debug the error. VM broke due to a full harddrive of error messages as far as I can tell. Error was somehow related to collision, from the little I could make out from the mass of error messages. Situation was a simple sphere dropping onto a surface.
| |
− |
| |
− | 10th : Wasted several hours getting a new VM working, had compile issues. Now working again, however. Came across an error in one of my functions whilst debugging, which involved NaN results. Working on finding the cause and will fix asap.
| |
− |
| |
− | 11th : Fixed NaN results, had a 0/0 division which I changed. Also refined code to better handle ground plane with respect to gravity from it (i.e ignore it). Gravity is still counted as 10 down, with the gravitational attraction seperate. Should be in position to finally make video tomorrow morning.
| |
− |
| |
− | 12th : Ran simulation and took raytraces at 50 step intervals. Looks like the groundplane moves, which isn't meant to happen. Will look into this and check it's not moving, and if it is stop it. Also changed use of gqa slightly, and implemented a, for now, debug usage rt, which I intend to make a user option later on, if they wish to automatically have the simulation saved as several (hundred/thousand) images.
| |
| | | |
| ===Development Schedule=== | | ===Development Schedule=== |