I am having a problem with individual planes becoming solid on the model i am working on. I can get one of the four radius’ to become solid but the other three have got me stumped. Any Ideas?
Without the SKP it’s impossible to say…
BUT let’s assume you geometry is very tiny…
I can see a very thin ‘sliver’…
Some of the facets might have Edges < ~1/1000" which means SketchUp can’t create them [its built-in tolerance].
If you Scale up x10 or 100 and do the processing then Scale back down again what happens ?
Tiny edges/facets can exist but they cannot be created…
In addition to what TIG suggested, please tell us what you started with and exactly what you did to make this happen. As he says, without a file and with a terse statement like that, we can only guess.
I would be glad to attach the SKP if that helps. I pretty new to this, I’m not sure how to scale. here is the file. Thank you so much for taking the time to help.
circuit breaker fuel tank 3.031.skp (25.8 KB)
Your model is not too small, so the problem must lie in your technique. I erased all except one end and pushpulled the remaining face back out, and got this:
Now all is clear.
it’s not an extrusion.
The two PushPull extruded ends work fine.
You can’t make quad faces for those ‘facets’, they need triangulating.
look at the ‘diagonals’ which have to be added by hand-stitching…
Or make a simple extrusion and Scale one end-face afterwards to make the taper…
1 Like
I did as slbaumgartner, and finally moved the edges at the end left, right and down to match the taper of the original.
Anssi
circuit breaker fuel tank 3.031.skp (44.3 KB)
1 Like
Ah TIG and Anssi, more observant than I was…I missed the taper! Sometimes perspective view tricks me!
One thing to note, the two end pieces are not perpendicular to the ground. They diverge.
I have triangulated in the past, but I was able to get one corner to fill in but the other three wouldnt go. This part tapers both ways, with a base that is parallel to the ground. That is what is really throwing me off. Any idea why one corner fills in and the other three wont? Is it because of the segmented radiuses? Any ideas on how to determine volume when it is finished?
Anssi,
It appears that the model that you created does not have triangulated facets? Can you explain the steps you took? Thank You Everyone for your input. Every time I use this program I learn something…
It is completely unintuitive, at least to me, but you can shift and scale the end of a cylinder in an amazing number of ways and the sides can still be handled by four-sided faces. So, since your shape is basically flat sides with sections of a cylinder at the corners, moving them around can often be done without needing triangles.
What Anssi did was to erase all but one end, extrude that face to the original length, then select first the top edge of the small end, including the adjacent round corners but not the sides, and move it down. Then similarly select one of the side edges and adjacent round corners and move it toward the middle. Finish by doing the same to the third side.
Checking the two faces one reports different number of edges, pp large on shows one face that does not form which allows correction in that area.
Deleted all connected edges, pp large rect to small location, selected the end face and scaled down to match small endcircuit breaker fuel tank 3.031_MAC1.skp (26.9 KB)
Note the small faces after pp of large rect showing issue with arcs which you need to correct 
If you form face on large rect at start of pp you can make your model a solid and it will then report vol of 3573.166958 in^3
Up date, correction of large rec.
It was also noted if one selects option to create face at start pp then makes coplanar faces and solid will not form without removing those. Correction of rec faces also changes vol rpt to 3575.30370in^3circuit breaker fuel tank _mac1a.skp (37.0 KB)
Sorry about being late…
As I said, I just selected the end edges and used the move tool to taper the shape. The faces stay quadrilateral because the edges stay parallel - the original rectangle is just distorted into a parallelogram.
Anssi