Angular and Rectilinear Acceleration checking

[Psychophanta]

As long as I know, today there is no way to achieve the Rectilinear (nor angular) acceleration of the mass center of a body.
I know, we can get its acceleration per tick (program loop), checking its instantaneous velocity in a loop, and then check it again for the next loop, and then substracting the current one from the previous one.

However, same as in the real world, a "static" body supported onto a surface, experiments an acceleration (which is reported by any acceleration sensor) which points colinear to the gravity flux; and, aparently, it is not moving to any direction.

So, is there any trick or so to "install" an accelerometer (acceleration sensor) to a EntityBody?
I mean, is there any way to know the acceleration which a body is experimenting?

Thanks in advance for your reading.

[DK_PETER]

You can get the LinearVelocity and AngularVelocity using...
https://www.purebasic.com/documentation ... ibute.html
Edit: Forgot to add..
You need to calculate the acceleration using the velocity values yourself.

[Psychophanta]

You can get the LinearVelocity and AngularVelocity using...
https://www.purebasic.com/documentation ... ibute.html
Edit: Forgot to add..
You need to calculate the acceleration using the velocity values yourself.

Yes dear, I already know that way. And it is what i am implementing. However, as stated, that way does not allow to know the gravity acceleration with experiments a body which is stored onto a flat floor, for example.
That's way my question and request for a #AngularAcceleration and #LinearAcceleration flags.

[DK_PETER]

Which is why, I wrote it as a confirmation to your question.
There is, at present, not any other way to do it.
But you could add it to the request and wishlist..

[Psychophanta]

Yes, thanks, but requesting almost never works

[marc_256]

Hallo Psychophanta,

- Do you want to read an real world accelerometer (acceleration sensor) ?
We (me and my daughter) are using it with arduino, and you can read the XYZ from the accelerometer via serial communication.

- If it is only in 3D game program software,
this is very simple if your object is not moving (static body)
X = 0.0
Y = 0.0
Z or gravity = 9.81 m/s2
but once, your objects is moving (dynamic body) this becomes very difficult.
You need to calculate/compare the old motion speed data with the new one in XYZ directions and rotations.
And delta linear acc XYZ and rotations must be calculated with the old and new speed, rotations and the time difference.
If your object is jumping from a height, gravity can be 0, or even negative ...
If your object is falling freely it will increase by about 9.81 meters per second every second...

I used this a long time ago, will see if I can find some info of my old program,
it must be stored on an old backup HDD somewhere.

Sorry if I did not understand your question very well.

marc,

[Psychophanta]

Hello Marc.

Yes, indeed you understood my question, and in 2016 i made several prototypes with the accelerometer well known MPU6050, or similars, which are extraordinary. In the 2001 I used ST and PIC micros, with industrial accelerometers from "Sensorex", and it was worse than MPU6050.
In 2016, with the MPU6050 and PB, I performed some tips to see in screen a cube and move it , and rotate it while i move and rotate the physical accelerometer.
This device just returns 2 vectors (6 scalar variables), one of them is the rectilinear acceleration and the other , the angular acceleration. Of course, relative to the chip itself.

Since i have other works on my desk for other stuff, I had to left this, but really is so interesting.
I am just trying to simulate my algos first with PB in 3D than plug an arduino and try. And this way is useful , upto certain point.

[Olliv]

Hello Psychophanta,

I doudt that the native velocity functions are already realistic... If yes, then the acceleration features should be added.