Unsolved Lower downforce with higher angle?

Hi, I have a triangle that adjust it's angle dynamically by the chassis movement, however I've found a strange behavior:

This triangle located at rear axle does not produce more downforce when having higher angle, even with stallAngle set to a very high value. Below two situations are having the same speed, same every other thing, but only with different ride height and triangle angles.

The one with 22 degree is producing 440N of downforce, while 31 degree is also producing 440N. I have noticed that in both situations, when you lift or lower the triangle angle, it will change base on this 440N, but 31 degree does not have more downforce than 22 degree. In the Jbeam my stallAngle is set to 0.5, which is close to 28 degree.

Code:

{"optional":true},
        {"liftCoef":80,"stallAngle":0.50}
        {"breakGroup":"rbm"},
        ["rh2r", "rdf", "rh2l"],
        {"breakGroup":""},
        {"dragCoef":0},
        {"liftCoef":0,"stallAngle":0},
        {"optional":false},

The most strange thing is that, on front axle it works as it should. As you could see, the front triangles, when at 27 degree, added with other parts it will produce 2200N of downforce, but when at 18 degree it will only produce 1900N of downforce. All other triangles are static. This means the angle change will have 300N of difference.

Code:

{"optional":true},
        {"dragCoef":15},
        {"liftCoef":55,"stallAngle":0.48}
        {"breakGroup":["fbm","lip_L","splitter_L"]},
        ["fh2r", "spl", "fh2l"],
        {"breakGroup":["fbm","lip_R","splitter_R"]},
        ["fh2r", "spr", "fh2l"],
        {"breakGroup":""},
        {"dragCoef":0},
        {"liftCoef":0,"stallAngle":0}
        {"optional":false}

Another thing to note is that, on front axle, the downforce nodes are connected to bumper, but on rear the node is directly connected to chassis panel, where refNodes also connectes to the same place.

So what should I do to make the triangle on my rear axle react more to triangle changes? Should I play with dragCoef and liftCoef?
--- Post updated ---
I have possiblly found the reason: The angle of attack on my front axle, is 5-15 degree with the higer ride height, and 25-30 degree with lower ride height. The rear axle however, they all exceed 25 degree.

It's like going from 0 degree to 15 degree has huge impact, while going from 25 degree to 35 degree there's very little difference.

I know it's not linear, but I didn't expect this much of a difference. I now get 200N+ difference when I adjusted nodes, position, without adjusting any those coef things. Before it was going from 20 degree to 35 degree, now it's going 5 degree to 20 degree, and it made a huge difference, while the angle difference remains 15 degree.
--- Post updated ---
Another thing discovered: strangely when I lowered stallAngle on front axle, I get more downforce, but the difference is smaller. Does this mean, actually the downforce is calculated with stallAngle, with stallAngle is the "maximum downforce angle", and it gradually decrease from the stallAngle, instead of decrease from something like 45 degree?

For example if i set my stallAngle to 0.45 which is about 25 degree, then at 25 degree I get 100% downforce, if I set stallAngle to 0.65, then I get less than 100% downforce when at 25 degree? I don't understand it.

If somebody can explain how this happened it would be super helpful.
1. Why do I get different downforce difference of high and low angle of attack, between front and rear axle?
2. Why do I get more downforce by decreasing stallAngle, instead of increasing it?