Explode function changes part size

If you are having problems with QCAD, post here. Please report bugs through our Bug Tracker instead.

Always attach your original DXF or DWG file and mentions your QCAD version and the platform you are on.

Moderator: andrew

Forum rules

Always indicate your operating system and QCAD version.

Attach drawing files and screenshots.

Post one question per topic.

Post Reply
2CV375
Newbie Member
Posts: 9
Joined: Tue Sep 05, 2017 11:36 am

Explode function changes part size

Post by 2CV375 » Wed Mar 25, 2020 12:08 pm

Hello,

I am using qcad/cam version 3.24.2.0 (3.24.2) on Windows 10.

The attached file TESTPART when imported into qcad and then exploded changes the size of the slots along the top of the drawing. I need to explode the drawing to convert it into a profile for cutting. If you watch the top left slot you will see it shrink as it is exploded.

To overcome the problem I exploded the part in nanocad and then further exploded the part in qcad. Is there any way of just using qcad?

Also, the circles in the testpart.dxf when exploded in qcad and then imported into my cam program disappear. THE same holes when first exploded in nanocad and then exploded in qcad still remain.

If anyone understands why this is happening, please let me know.

I think the original part was developed in Microstation.

I am happy to continue to use my current solution if that is the best way.

Thanks,

Daniel.
Attachments
TESTPART.dxf
(133.79 KiB) Downloaded 475 times
TESTPART-NanoCadQC.dxf
(267.61 KiB) Downloaded 469 times
TESTPART-NanoCad.dxf
(317.32 KiB) Downloaded 481 times

CVH
Premier Member
Posts: 3416
Joined: Wed Sep 27, 2017 4:17 pm

Re: Explode function changes part size

Post by CVH » Wed Mar 25, 2020 1:49 pm

Loaded the File
Select All
Explode (for the Blocks)
Explode (for the resulting complex entities)

At first glance all looks fine but the exploded entities get the parent linetypes among wich some DNG STYLE ..
I see also that this way there are duplicates or about duplicates.

Found block *U2 and *U3 on top of each other.
Angled at 269,9843° ??
Probably I exploded too many things.
What is the Layer combination with the actual selection to explode?
Block *U4 has 2 hatches and looks crooked. (most do)

And indeed:
A solid hatch in a Block is exploded as a smaller size.
Could be a BUG.
Hint: Explode the solid Hatch in the Block definition and you get the boundary.
A boundary, what are simple entities, explode correct.

I do not understand the hatch without a boundary.
The hatch is only the fill and needed to give that space another representation.
E.g. a hole or slot painted black.
You still would need a boundary or the actual outline of the hole/slot.
I am not sure that the geometry of a hatched surface is 100% preseved in a hatch.
Going back it can get even worser.

Hatches from a Circle as boundary are exploded to 360 degree Arcs.
That could be an issue. Some soft don't accept this but usually they support two 180 degree Arcs.
Reversed/ Not Reversed can be an issue.
Another point is that those arcs have lineweigth 0.00mm.
This lineweigth is included for special needs in some target application.
'Default' lineweigth could be an issue too. Try 0.05mm.

Regards,
CVH
Last edited by CVH on Wed Mar 25, 2020 2:06 pm, edited 1 time in total.

CVH
Premier Member
Posts: 3416
Joined: Wed Sep 27, 2017 4:17 pm

Re: Explode function changes part size

Post by CVH » Wed Mar 25, 2020 1:54 pm

Are you sure about the placement of those slots?
I am sure that not all centerlines are at the center.
CVH

2CV375
Newbie Member
Posts: 9
Joined: Tue Sep 05, 2017 11:36 am

Re: Explode function changes part size

Post by 2CV375 » Wed Mar 25, 2020 9:47 pm

Thank you.

Yes, I agree the slot placement is incorrect. The dimensioned drawing I have as a PDF for checking suggests a shift of about 1,5mm (!). Additionally, the imported dimensions are different by a few tenths of a mm. A 0,25mm difference is tolerable; however, it would be good to understand why. The shift I think is an error in the export as when working with these parts they have always been incorrect by a little bit.

These inaccuracies appear to stem from the original application extraction and how the part is represented in that application. I do believe the original design was done in Microstation and it represents a hatch differently to Autocad, perhaps more so as a DXF.

Yes, the 2 x 180 degree arcs work (as I split the circle). The "however" is that when I use nanocad, the circle is converted into a format that is understood by the other cad package. It is not split.

At the moment I do have a solution. It would be good if there is a bug, that it is understood or fixed for others.

CVH
Premier Member
Posts: 3416
Joined: Wed Sep 27, 2017 4:17 pm

Re: Explode function changes part size

Post by CVH » Wed Mar 25, 2020 10:58 pm

Like in your other topic:
The two semicircle thing has no solution...YET
It would not be to hard to code a script for that.

Why a hatch based on a circle has a double semicircle explosion:
A boundary is stored quite similar as a Poly.
Catch-22 is that a bulge can not represent a full circle.
This is the limit of the function being infinite.

It could be that 'the others' have algoritms in place that gaurd for this.
Auto-replacements have their downsides too.
As most auto triggered actions do.

I do prefer an action triggered by the user.

Regards,
CVH

CVH
Premier Member
Posts: 3416
Joined: Wed Sep 27, 2017 4:17 pm

Re: Explode function changes part size

Post by CVH » Thu Mar 26, 2020 8:20 am

2CV375 wrote:
Wed Mar 25, 2020 9:47 pm
Yes, the 2 x 180 degree arcs work (as I split the circle).
The "however" is that when I use nanocad, the circle is converted into a format that is understood by the other cad package.
It is not split
This confuses me.
Do you need 2 semicircle Arcs, 1 full Arc or 1 Circle?

TESTPART-NanoCad.dxf looks similar as TESTPART.dxf, both have a circular Hatch in a Block called HS0002_1.
in TESTPART-NanoCadQC.dxf these are full Arcs.
Exactly the same as Qcad would explode the circular Hatches.
Ok, almost exact: Start and end angle:
Qcad 0° vs NanoCadQC 0.028584°


The odd numbers should trigger the awareness that there is something wrong with the design.

Lets Break it down:
Added Layers: _Centerlines ; _Hatch ; _Boundary
It is handy to diversify Block elements on dedicated layers.
In this case hiding layers _Hatch & _Centerlines will only display the geometry.

Edit Block HS0002_1
Moved the 2 Lines and the Hatch ; Cloned the Hatch in _Boundary and exploded it.
Apart from the radius all properties look ok. R=9
You could replace the boundary by a circle R=9 if preffered.
This Block is used 4 times scale 1, angled 0.028584°.

Edit Block *U1
Moved the 3 Lines and the Hatch ; Cloned the Hatch in _Boundary and exploded it.
Here the placement and the properties DO NOT look ok.
Most likely this should be a 26x18 slot R=9. Not hard to do that overnew.
This Block is only used once at scale 1.222, angled 269.9843°.

Edit Block *U2 and move entities.
The placement and the properties DO NOT look ok.
Slot 28x18 R9 slot.
This Block is used 5 times at scale 1, angled 269.9843°.

Edit Block *U3 and move entities.
The placement and the properties DO NOT look ok.
This is a 18 wide slot that is 26.183307 long R=9, only .183307 longer as *U1
This Block is used 7 times at scale 1.222, angled 269.9843°.
!! 5 times in conjunction with *U2.

Edit Block *U4.
What appear to be 5 Lines are in fact 8 Lines, 2 are duplicates, 1 is almost.
Keeping only 5.
There are 2 duplicate Hatches, exploding 1 in _Boundary
As with the others the placement and the properties DO NOT look ok.
Most likely this should be a 70x24 slot R=12.
This Block is used 20 times at scale 1.09, angled 0.0°.

All Blocks have origin (0,0), one of the few things that looks evident.
Apart from that: the positioning of the blocks DOES NOT look ok and isn't simple.
The workpiece contour sits on 3 Layers with different Weights.
All is slanted and right, high there is an orphan.

Most CAMs will choke on the file as is.
You got some work in front of you.

Regards,
CVH

2CV375
Newbie Member
Posts: 9
Joined: Tue Sep 05, 2017 11:36 am

Re: Explode function changes part size

Post by 2CV375 » Thu Mar 26, 2020 12:53 pm

This confuses me.
Do you need 2 semicircle Arcs, 1 full Arc or 1 Circle?
Sorry for the confusion and I appreciate the detail. I will analyse what you have provided tomorrow evening.

I need 1 full arc; however, one full arc does not always work. Yes, confusing.

I have included four files.

(1) Split circles load. This file loads into the cam program. I split the "circle" into two half circles. This file is "Circles_No_Load" + ms with 2 parts.
(2) Circles no load. This files is missing the circles when it loads into the cam program. This is an explode and clean up in qcad only.
(3) Circles load no split clean. This is the dxf that is output from the cam program after loading the part and automatically purging rubbish entities. It is file (4) loaded and saved in the cam program.
(4) Circled Load no split. On this file I performed an explode in nanocad I then cleaned the file up in qcad.

Thanks for your help CVH.
Attachments
Split_Circles_Load.dxf
(174.66 KiB) Downloaded 500 times
Circles_No_Load.dxf
(173.67 KiB) Downloaded 459 times
Circles_Load_No_Split_Clean.DXF
(17.25 KiB) Downloaded 460 times
Circles_Load_No_Split.dxf
(189.2 KiB) Downloaded 449 times

CVH
Premier Member
Posts: 3416
Joined: Wed Sep 27, 2017 4:17 pm

Re: Explode function changes part size

Post by CVH » Thu Mar 26, 2020 2:07 pm

(1) Split circles load.
2 semicircles Arcs 0-180° & 180-0° Non-Reversed.
Linetype: DNG Style 4, Lineweight: 0.00mm

(2) Circles no load.
1 full Arc 0-0°Non-Reversed.
Linetype: DNG Style 4, Lineweight: 0.00mm

(3) Circles load no split clean.
1 Circle
Linetype: By Layer (Continuous), Lineweight: By Layer (Default)

(4) Circled Load no split.
1 full Arc 0.028584-0.028584° Non-Reversed.
Linetype: DNG Style 4, Lineweight: 0.00mm

Summary
What doesn't fail:
- Linetype: DNG Style 4; Continuous
- Lineweight: 0.00mm; Default
- Type: Circle; semicircles; full Arc 0.028584-0.028584°
Typical troublesome in bold , here they aren't

What fails:
-full Arc 0-0°

Lets try to alter only one full Arc to 0-0° in (4)..... :wink:
And see if the claim holds.

Regards,
CVH

2CV375
Newbie Member
Posts: 9
Joined: Tue Sep 05, 2017 11:36 am

Re: Explode function changes part size

Post by 2CV375 » Thu Mar 26, 2020 8:36 pm

What fails:
-full Arc 0-0°
Thank you. It makes sense that a full Arc 0-0°does not load if the application is only expecting entities that follow a path (0-360). I do accept that a 0 and 360 are the same point. QCAD is quite tolerant. :-)

CVH
Premier Member
Posts: 3416
Joined: Wed Sep 27, 2017 4:17 pm

Re: Explode function changes part size

Post by CVH » Fri Mar 27, 2020 4:54 am

And there is a funny thing about it.
Had a discussion with Andrew a while ago.
In a way I expected issues similar as this.

Draw 2 Arcs CCW. Placement and Props don't matter a lot.
1) set Start Angle 0; set End Angle 0 >> Sweep = 360
2) set Start 0; set Sweep 360 >> End Angle = 360

Something similar goes for CW Arcs, negative sweeps.
Same goes as we use math to alter the angles.
https://www.ribbonsoft.com/bugtracker/i ... sk_id=1924

IMHO Normalisation is needed.

Hint: Try to force the 0-360 notation on one off your full arcs ....

CVH
Last edited by CVH on Sun Dec 20, 2020 7:26 am, edited 1 time in total.

CVH
Premier Member
Posts: 3416
Joined: Wed Sep 27, 2017 4:17 pm

Re: Explode function changes part size

Post by CVH » Fri Mar 27, 2020 6:59 am

Coming back to the topic.

Confirmed:
Negative Scales (flipping) and Hatches in Blocks have an issue exploding.

Bugreport at:
https://www.ribbonsoft.com/bugtracker/i ... sk_id=2059

Regards,
CVH
Attachments
HatchInBlockScaleErr.dxf
(107.42 KiB) Downloaded 471 times

User avatar
Husky
Moderator/Drawing Help/Testing
Posts: 4935
Joined: Wed May 11, 2011 9:25 am
Location: USA

Re: Explode function changes part size

Post by Husky » Fri Mar 27, 2020 9:05 am

Hi Daniel,
2CV375 wrote:
Wed Mar 25, 2020 12:08 pm
The attached file TESTPART when imported into qcad and then exploded changes the size of the slots along the top of the drawing. I need to explode the drawing to convert it into a profile for cutting. If you watch the top left slot you will see it shrink as it is exploded.

To overcome the problem I exploded the part in nanocad and then further exploded the part in qcad. Is there any way of just using qcad?
No - not to my knowledge. TESTPART has 3D information. QCAD is a plain 2D program and 3D information in a drawing is a recipe for trouble. :wink:
2CV375 wrote:
Wed Mar 25, 2020 12:08 pm
I am happy to continue to use my current solution if that is the best way.
Most likely that is the best way. Without going to deep into it everything in the TESTPART-NanoCad.dxf drawing seems to be a usable basic. Load the drawing, purge the unused blocks (Menu/Block) and if needed use the main drawing to create new blocks with valid names.
Work smart, not hard: QCad Pro
Win10/64, QcadPro, QcadCam version: Current.
If a thread is considered as "solved" please change the title of the first post to "[solved] Title..."

CVH
Premier Member
Posts: 3416
Joined: Wed Sep 27, 2017 4:17 pm

Re: Explode function changes part size

Post by CVH » Fri Mar 27, 2020 9:25 am

Husky wrote:
Fri Mar 27, 2020 9:05 am
TESTPART has 3D information
Could you point me to these Z isnot zero Props because I have missed them.

Thanks,
CVH

CVH
Premier Member
Posts: 3416
Joined: Wed Sep 27, 2017 4:17 pm

Re: Explode function changes part size

Post by CVH » Fri Mar 27, 2020 9:37 am

Husky, There are no Z <> 0
Let it be.
The bug is fixed and closed already.
CVH

Post Reply

Return to “QCAD Troubleshooting and Problems”