Solid Sweep with Equation Curve

[ilyas_drawbridge]

Hi Everyone,

I'm trying to perform a solid sweep cut via an equation curve, and to get the curve smooth, I want to use a harmonic function around theta. However, the solid sweep does not seem to compute with it. I tried cycloidal and parabolic as well, and they don't seem to compute. But for linear function, it works. Attached is the sample file with the harmonic function.

Harmonic:90*0.5*(1-cos(180*t))

Parabolic:90*(2*t^2) for [0-0.5] and  90*(1-2*(1-t)^2) for [0.5-1] 

Linear:90*t

Any thoughts?

[ilyas_adnm]

Update: The sweep seems to compute for 0<t<1, but not compute when it is equal to 0 or 1. I guess this is the critical limit where dtheta/dt =0. Any thoughts on how I could address this?

[karthur1]

I was not able to make the sweep work either, but look at your equation.  Are the units what you expect for the f(t)?

[ilyas_drawbridge]

Thanks @karthur1 for looking into this,

t would represent the full cycle of my sweep. i.e.: sweep starts at t=0 and ends at t=1. I intend for one cycle to be 90 deg rotation and 50 mm z translation. Hence, unitless would be the correct unit for t.

for r(t) and z(t), both are correct as per your screenshot.

for the theta(t)= the units assigned by the system seems to be overfactored, but the overall product of the units are correct. Ideally it should be: 90 deg * 0.5 ul * (1 ul - cos (180 deg * t))

I'm currently looking at parabolic equation to define my sweep. It seems to work for t not close to 0 or 1. My current plan is to use make the sweep as wide  ranging as I can, and then blend the ends using a G2 fillet. 

Appreciate any thoughts about this.

Regards

Ilyas

[johnsonshiue · Accepted solution]

Hi! Indeed, the equation curve at both ends have extreme high curvature that exceeds the modeling tolerance (10E-5mm). Inventor cannot create such geometry. Change the parameter range from 0~1 to 0.0001~0.9999. Then it should work.

Many thanks!