Boolean split fails? (Attempting to create a mold)

I am new to Rhino and CAM and am attempting to create a mold. I have used Rhino with Tsplines to model a positive object that I wish to create from the mold. (The object is an aerodynamic “power pod” for a powered glider, essentially a nacelle on an airfoil-shaped post). I now need to model the “mold” or negative around the object. The mold will be used to mold the actual object in carbon fiber.

I have done the first step successfully – I modeled a box around the object, then did a Boolean Difference, which produced a valid polysurface – basically the mold which needs to be cut into halves.

However, I have created a cutplane (two split the mold into left and right mirror image halves), and when I attempt to do a Boolean Split between the box and the cutplane, the split fails.

As far as I can tell, the intersection is valid. What do I need to do?

Rickpower pod mold - narrow version cutplane.3dm (444.0 KB)

You can try the Split command instead of boolean in this case.
Select the item you want to split. In this case the mold,
type run split command then the object to split with then finish.

hmm… sorry the split didn’t work well…

The Split command works but does not produce the correct results. For some reason it leaves a thin shell of material around the original positive, but the rest of what should be solid material is not – the rest of the box becomes totally hollowed out. It’s weird, and does not give the results I need. Thanks.

The naked edge inside your model is causing the problems.

Someones else might be able to show you how to fix this but to get it to split and prove it was the problem I just sliced off a tiny piece of the end to remove the point and capped. It split fine then.

Probably quicker to fix the topology of your T-splines model than trying to fix this in Rhino.

sorry about that.
I also noticed the naked edge after checking…

I assumed it was a revolved surface, so just cut off the problematic surface…
First extracting a isocurve from the original surface by extractisocurve, then trimmed the
problematic surface with trim and also trimmed the isocurve extracted.
Then revolved around a center point.

Put the parts back in place.

I tried to do boolean, but was asking for more tolerance, so did a split surface then cap.

I also would like to know how others would go about fixing the point naked edge…

power pod mold - narrow version cutplane_2.3dm (902.8 KB)

I couldn’t figure out how to remove the tip and cap it? Or how I would otherwise fix this naked edge (which is just a point?) But your solution gave me another idea – if I move the cutplane just a few thousands off center, which will be just fine for my purposes, the boolean split worked fabulously. Thanks!

Edit: Just saw Toshiaki’s answer and yes, if I can fix that naked edge/point, would love to know how!

offset boolean is ok, but you’ll still be left with the naked edge on one side…
well, I guess mirroring the side which doesn’t have the naked edge can be solution in that case…

I just took a look at that file you uploaded, Toshiaki. Very nice work and I think it solved the problem beautifully! I actually realized after the fact that my offset solution would be very bad at the trailing edge of the airfoil on the lower post, so splitting right down the middle is really much preferred, which is what you’ve managed. Thanks and this is what I can use moving forward!!


@rjtw It was not really a solution just pointing out what the problem was…that is… not Rhino but your modelling in t-splines. As I wrote previously this would probably be better fixed in t-splines than Rhino especially if you want to make changes/revisions in t-splines later and re convert to nurbs.

I always check my t-spline objects after I convert them to nurbs. If you had in this case then you would have seen the problem and gone back to t-splines to fix it.

If you want to use t-splines in the future and learn how to not run into this problem again …or at least how to fix it then I would suggest you post the t-splines model and let some of the t-splines guys give you some suggestions on how to avoid this problem in the future.

Glad you got it sorted in Rhino anyway.

Could you create a pre-split solid to subtract your shape from then just do a Boolean difference on both pieces? Then you wouldn’t have to split it after the Boolean.