Private Sub Timer1_Timer()
For i = 1 To nodes
    'reset forces
    n(i).Fy = 0
    n(i).Fx = 0
Next i

'take care of gravity
For i = 1 To nodes
    n(i).Fy = n(i).Fy + n(i).m * g
Next i

'take care of the oscilations 
For i = 0 To UBound(t)
With t(i)
    d = Sqr((n(.front).x - n(.end).x) ^ 2 + (n(.front).y - n(.end).y) ^ 2)
    
    x = d - .space 'displacement from equilibrium position

    'bunch of calculations that handle forces due to springs. Im quite confident 
    that these work. However, if x is not 0 when it should be, the program will
    calculate forces that shouldnt be there, and then i get acceleration that
    shouldnt be there, and the program blows up. And the 1.7 E-15 builds up
    really fast....

    F = .k * x
    n(.front).Fx = n(.front).Fx + F * (n(.end).x - n(.front).x) * d
    n(.front).Fy = n(.front).Fy + F * (n(.end).y - n(.front).y) * d
    n(.end).Fx = n(.end).Fx - F * (n(.end).x - n(.front).x) * d
    n(.end).Fy = n(.end).Fy - F * (n(.end).y - n(.front).y) * d
    
End With
Next i

'calculate acceleration
For i = 1 To nodes
With n(i)
    aX = .Fx / .m
    aY = .Fy / .m
    
    .vX = .vX + aX
    .vY = .vY + aY
End With
Next i

'move particles
For i = 1 To nodes
With n(i)
    .x = .x + .vX
    .y = .y + .vY
End With
Next i

'
    dblnx = n(.front).x - n(.end).x 'is this backward?
    dblny = n(.front).y - n(.end).y ' "                  "
    d = dblnx * dblnx + dblny * dblny 'minimizes errors due to Sqr and ^ 
    
    x = d - .space * .space 'rewrite so that .space is already stored as .space^2 to speed things up and remove a small error.

    F = .k * x
    d = Sqr(d) 'defer the Sqr until absolutely needed
    n(.front).Fx = n(.front).Fx + F * (n(.end).x - n(.front).x) * d
    n(.front).Fy = n(.front).Fy + F * (n(.end).y - n(.front).y) * d
    n(.end).Fx = n(.end).Fx - F * (n(.end).x - n(.front).x) * d
    n(.end).Fy = n(.end).Fy - F * (n(.end).y - n(.front).y) * d