Son of Panel Fit 250812

[pittsallen]

Hello Fusion forum,

Working with the excellent input from jhackney
Chandelier Assy - JRH.f3d, attached
some additions were made to have a complete
chandelier assembly:

When the assembly was disassembled and made
into separate parts for printing it
was discovered that the top part of
the six panels is too big for the Bambu
Carbon One printer. The printer plate
had been measured at 254 mm (ten inches) but
the printer will not print larger than
229 mm (9 inches).

Back to the drawing board.

The panels were reworked to a smaller
size.
chandelier_assembly_250811.sfd, attached

Struggling with how to fit the panels together
so they make a hexagon on the top
and a hexagon on the bottom as shown in
Chandelier Assy - JRH.f3d 

The angle for the joint at the top and the
bottom has not been set for the mitered joint and
key block. The plan is to, after
the panels are set use the assembly
to discover the angles for the
faces where the mitered key joints
come together.

Studied jhackney's solution to figure
out how it was solved. The timeline
suggests that a circular pattern
was used. Not exactly sure how this was
done and some hours of study
and experimentation have not led
to a solution.

Also tried more than several combinations
of Align and Joint. It is thought
that making a joint between the top
and bottom panels before Aligning or
Jointing keeps the top and bottom
moving together.

Any ideas on how to approach the compound
miter are appreciated.

Thanks.

Allen

[jhackney1972]

You have a design problem in the lower half of your panel.  It is not symmetrical so I cannot achieve a clean joint on the bottom of the assembly.  I have included my beginning assembly layout, where you can see the problem in the bottom joint, and a model which matches up two panels to show the symmetry problem.  When you solve the symmetry problem on the lower section, let me know and I will finish the assembly.

[TrippyLighting]

These problems are encountered because the underlying sketches are inadequate.

None of the objects in the first sketch is fully dimensioned and constrained. That makes the design very ... fragile.

A lot of sketch work and debugging can be eliminated by sketching only half of the object and then mirroring the 3D geometry. 

[pittsallen]

Hello @jhackney1972   and  @TrippyLighting,

The bottom panel is redrafted with special care to keep the object symmetrical.

Thanks to Mr. Hackney for the tip on checking for symmetricity.

To Trippy's point, a great deal of time and effort can be obviated by drawing half of a symmetrical object and using the Mirror command.
An  example is the lowest curve and the tracery in the bottom panel. 
If a second Mirror is then used to get a geometry in another part of the sketch and if the curve is picked up for the second mirror
the second mirrored curve is sometimes slightly offset from the first.
Ever had that problem?
Perhaps the issue is with choosing the Mirror axis.

Thanks.

Allen

 

[jhackney1972]

Your bottom section still has some symmetry problems.  I will let the video explain.  Because of this symmetry problem the joint at the bottom will not be very pretty, as you can see.  I am using the top part of the panel for my alignment in the assembly.  It is not perfect but pretty close from side to side as shown in the screen capture.  The dimension I am showing is the vertical distance between each top common point.  The bottom is way off as shown in the video.  Models attached.

[TrippyLighting]

---

@pittsallen wrote:

...

Ever had that problem?
Perhaps the issue is with choosing the Mirror axis.
 

---

No, I rarely ever use mirrors in sketches and when I do, it is for less than a handful of sketch objects.

Mirroring an entire sketch half is not a recommended technique.

One of the issues in this design's sketches, as I've already mentioned, is that they are not properly dimensioned and constrained.

I believe many of the splines in your first sketch can be replaced by arcs, which then can and should be fully constrained, dimensioned.

Many of the splines also have more spline control points than needed.

Additionally, this is very Fusion-specific, and it appears that many users are unaware that mirroring Fit Point Splines is not a recommended workflow. The expectation is, of course, that when a fit point spline on one side is edited, the mirrored fit point spline is an exact copy. However, that isn't always the case. Edits can to fit point splines can destroy the symmetry.

That is another reason why mirroring actual 3D geometry is preferable to mirroring sketch.

The general recommendation for symmetric objects is to sketch half of the object, then, in this case, extrude that half and then mirror the geometry. 

I suggest that you follow and implement this advice and post intermediate steps so we can provide proper guidance.

[TrippyLighting]

I had overlooked the most obvious problem.

The sketch doesn't even use the sketch origin for symmetry. 

I would start this design completely from scratch!!!

[laughingcreek]

You're getting some really good advice here, and things will definitely get easier if you start to follow it.  One more suggestion I would make is to go ahead and create a plane that is set at the correct location for the profile sketch.  (this technique would be what's referred to as "building in place").  no joints or moving things into place necessary. it will make the rest of the modeling tasks soooo much easier to deal with.  changes/edits will be easier to make also.  attached is a simplified example of doing that.  I'd suggest going thru the timeline till you understand what is happening here.

cheers.

[John_Wright]

I agree entirly with @laughingcreek as to doing this with "builiding in place"

This solve all of your issues with caluclated angles, alighnment, joints etc. It is how I would approach it.

Don't get stuck in the "sunk cost fallacy" with your model. It really would work out better if you started from scratch.

[pittsallen]

Hello @jhackney1972  @TrippyLighting @John_Wright  @laughingcreek and the Fusion forum,

Have reworked the bottom panel to eliminate the symmetry issue. Found two
points on the lower right side where there should be one point that
were fouling the works.

Using jhackney's diagonal measure check the panel in attached file
chandelier_panel_250813.f3d
the diagonals, which were 3 mm off are now .8 mm different. While

not perfect, it is hoped that that is close enough to make a better
fit of the miters on the bottom.
(jhackney, how do you get two measurements to show at the same time?)

On starting over from scratch: Given the time and effort invested
there is certainly a willingness to start over to improve the model.
The desire to see if the present model can be tweaked into an acceptable
solution is motivated by the uncertainty that given the complexity
and number of processes required to complete the assembly
there is an excellent probability of two possible outcomes:
1. The completed restart is no more perfect than the existing model.
2. The completed restart is worse than the existing model.

The desire for perfection is understood and admired. An aspect
of building the chandelier that has an impact on the acceptable
level of perfection is the 3D printing process. Many times
a model much less complex than the chandelier has been modeled
in Fusion with precision exactness. But when printed by the
Bambu Carbon One the limitations of the printer obviate
the precision.

Thanks.

Allen

[TrippyLighting]

---

@pittsallen wrote:

...

On starting over from scratch: Given the time and effort invested
there is certainly a willingness to start over to improve the model.
The desire to see if the present model can be tweaked into an acceptable
solution is motivated by the uncertainty that given the complexity
and number of processes required to complete the assembly
there is an excellent probability of two possible outcomes:
1. The completed restart is no more perfect than the existing model.
2. The completed restart is worse than the existing model.

---

If you follow the instructions given so far, a model that follows best practices will be far superior, to your current model because the model will be more "perfect" than the current approach.

The goal here is not just to make this model better, but to teach best practices., so these problems can be avoided in future projects!

[jhackney1972]

Well you modified the lower portion of the panel to get a descent joint at the bottom but that joint cuts into the feature detail.  I think you still need some modification.  The video will help you understand.  Model is attached.

[pittsallen]

Hello @jhackney1972  @TrippyLighting @John_Wright  @laughingcreek and the Fusion forum,

Reworked the tracery on the bottom panel so the miter does not interfere.

Then piggybacked on Mr. Hackney's latest Chandelier Assy -JRH.f3d and
substituted the reworked bottom panel using Align.
And I managed to cut the right side miter.

But now I am stuck on the left side miter. 
I could create a plane on the right side to

draft a sketch to cut the miter. But when trying to create a plane for the left side miter
some difficulty is encountered finding a reference from which to create the plane.

The reworked panel is attached:
chandelier_panel_250813_b v2.f3d

Chandelier Assbly 250813 v2.v2.f3d
is my attempt at mitering the bottom panel.

I think I am pretty close if I can get the
left bottom panel mitered. Any adult
supervision would be greatly appreciated.

Thanks.

Allen

[jhackney1972 · Accepted solution]

Your assembly is attached.  The Miters look good and they do not interfere with the panel features.  I have attached a single panel also.  I did not do the miter key as I need you sign-off on this assembly before I do anymore.  You of course can add them yourself.  If you have me do them, please tell me the approximate size you desire.  Your model still has "color" in the timeline you see to be ignoring.

If this answers your question, please select the "Accept Solution" on my post. If you need further help, please ask.

[pittsallen]

Hello @jhackney1972  @TrippyLighting @John_Wright  @laughingcreek and the Fusion forum,

Pheidippides ran twenty-six miles from the plains of Marathon to the city of Athens
uttered one Greek  word and dropped dead. The word was 'nike' or in English 'victory'.

 

A substantial nike (not the tennis shoe kind) was acquired today in getting the panels
to print.
The completed chandelier will require twelve panels, six
top and six bottom, six legs the centerpiece and the rosette.
Each panel takes about two and one half hours to print.

Thanks to the four gentlemen, and especially Mr. Hackney for their excellent counsel and advice.

Will post photo when panels joined.

Allen 

[TheCADWhisperer]

@pittsallen 

Most of the arcs are not Tangent and the entire design process could be simplified.

Note that the resulting geometry is not symmetrical.

[pittsallen]

Hello @jhackney1972  @TrippyLighting @laughingcreek  @John_Wright  @TheCADWhisperer and the Fusion forum,

The input from the forum community is much appreciated. Several times in designing the
chandelier a problem that put the design in a rut from which there was no way to break free
was solved by the altruism and expertise of the gentlemen in the Fusion forum.

So, a feeling of selfishness occurs in that time and proficiency were received without
reciprocity.

Perhaps rectification could be accomplished by demonstrating the fruit of our labors.

Six individual panels were put together from twelve top and bottom pieces as shown in
forum thread 'Son of Panel Fit 250812' post #15.

The challenge then was to fit the six panels together in a regular hexagon,
The usual process when supergluing 3D printed ABS plastic pieces is to dry
fit to check for correctness, place the adhesive and clamp.

But the dry fit was problematic because there is no way to clamp two pieces at a
60-degree angle. Tried rubber bands around the middle with no luck. Considered
building a hex rig for clamping.

Finally, after several hours of struggling it was decided to glue them one at time
without a dry fit of the complete object and rely on the 60-degree bevel to
get a regular hexagon at the top and bottom edges

I must be living right because it worked and when the sixth panel was set in place
it came together correctly.

The final quandary was getting the legs to fit to the centerpiece.

Again there
was no way to fit all the legs to the centerpiece for a dry fit and blind
faith was the only option.

Again my luck and the Fusion design held up and the complete chandelier appeared.

The next step is to place the LED COB strips.

When the LED strips are set the strips will be covered with painters' tape
and the chandelier will be painted brown to match the Queen Anne
dining room table and chairs.

Allen Pitts