+- Wings 3D Development Forum (https://www.wings3d.com/forum)
+-- Forum: Wings 3D (https://www.wings3d.com/forum/forumdisplay.php?fid=1)
+--- Forum: Interface & Usage (https://www.wings3d.com/forum/forumdisplay.php?fid=3)
+--- Thread: Measurements - Need some help (/showthread.php?tid=86)
RE: Measurements - Need some help - Fonte Boa - 12-06-2012
Until the following point i came (cant record video using mrs Fonte Boa notebook):
![[Image: wingstrigonometry.jpg]](http://imageshack.us/a/img62/269/wingstrigonometry.jpg)
Therefore if i use Shift|Planar on the 2 selected faces at pic (one is visible in red, the other is paralel to this first one), picking any vertical edge as Axis and cilinder circular face as center, the transversal piece top and botton faces slide in direction of cilinder...
RE: Measurements - Need some help - Fonte Boa - 12-07-2012
For God sake, Micheus!
It is impressive how exact it can be if math calcs are involved this way.
Cant do that without build a formula for each case, sorry: who knows one day i'll be studying this with my son?
Congratulations, man.
RE: Measurements - Need some help - puzzledpaul - 12-07-2012
(12-07-2012, 12:16 AM)Fonte Boa Wrote: It is impressive how exact it can be if math calcs are involved this way.
Totally agree (note my post 25) ... trig's good stuff
Luiz
Re shift approach
Cylinder's the wrong diameter I think ... looks like 1 unit rather than 2 (should be same diameter as width of columns)
The transverse piece (as you call it) - should be bridged between the top of the left column and the bottom of the right column ... not a separate object. (at this stage it should be one object, not 3)
The elements to be shifted are 2 edges - not faces.
These edges are where the transverse is bridged to the columns ... the 2 short edges inside the (acute angled) corners (joining the front to the rear of the N) ... a bit like an 'armpit cavity'
When you use shift planar > Y, these 2 edges will move in opposite directions, one up one column, one down the other column ... depending on mouse drag.
Land of zzzzz time here - if still not got it overnight, I'll do some grabs
pp
RE: Measurements - Need some help - puzzledpaul - 12-07-2012
Any use / help?
2 different approaches
![[Image: Ncombo.gif]](http://smg.photobucket.com/albums/v492/pppix/Ncombo.gif)
1 basic N with bridged diagonal / transversal ... + ref circle (extracted end face) + edges to be shifted
2 Shift | Planar > LMB ... select anything that gives correct Y
3 select centre of circle (or cold be verts at ends of edges being shifted) to provide operation origin / datum
4 Mess around until diagonal edges tangenial (zoom in for this)
5 create block of overall correct external dimensions, place edges as necessary to define columns, connect diagonal across square
6 Cut N ... I used 10 here ... more, better but less clear in tut.
Vert | deform inflate > RMB ... select vertshown for centre
7 select vert for rad
8 after dragging 100% (hold shift to constrain)
9 Tools | Connect to draw tangent. Place end point / vert on edge past rad, drag until line tangential.
Rinse, repeat or use edge as reference for another parallel.
Tidy up and shell extrude required shape ... or finish off the wine if any left from previous post.
(needs some different cols - done in hurry ...)
pp
RE: Measurements - Need some help - Fonte Boa - 12-07-2012
Hi, Paul: good morning (here).
Shift with the sphere solved the problem via eyeball.
Therefore, after Bridge, i selected transverse faces, hit L, then Loop Cut and created an apart object. Go to Body Mode and Invert. Select its longest edges and use Intersect. The rest you already know.
The video:
http://www.youtube.com/watch?v=CSCVJEeQRvY&feature=youtu.be
Which commands did you use in the 2 additional approaches, Paul (specially the 2nd one)?
Why dont you move this thread to "Interface and Usage"?
RE: Measurements - Need some help - puzzledpaul - 12-07-2012
Luiz - each to their own, but I tend to view in w/frame mode when doing this sort of eyeballing. Sometimes also use hard edges on ref objects, so's the different colour shows up better.
(probably be out soon with my brownie box cam as light looks promising - so don't expect quick responses)
pp
RE: Measurements - Need some help - Geta-Ve - 12-07-2012
Just wanted to quickly respond and say that you guys are wonderful! A lot of techniques to study here. I think once I get a couple bucks in my pocket (after Christmas probably) I am going to pick myself up a Trig for Dummies book and go through the whole thing. As well as hop on over to Khan Academy.
If there are any good resources you guys know of, off the top of your heads, I wouldn't mind the nods.
I will properly respond later to each of you. But thank you so much!!
RE: Measurements - Need some help - Fonte Boa - 12-07-2012
There are many trig lessons on youtube, Mike, and probably you'll find many stuff about using google: good luck.
RE: Measurements - Need some help - puzzledpaul - 12-08-2012
Whilst I don't realistically imagine this will be of use to anyone, it might just provide some solace for insomniacs
![[Image: Ntutdrawcombo5_01.gif]](http://img.photobucket.com/albums/v492/pppix/Ntutdrawcombo5_01.gif)
1. Standard start - 2 columns, LH one duped for the diagonal and rotated a little at random to show it exists.
To align the upper face of this diagonal with the target vert, select whole object, Rotate > Rot to target > MMB ... and select edge shown (in yellow) as rotate axis.
2. Select same edge to provide origin for rotate axis.
3. Select vert shown as the point to be aligned
4. Select the 3 verts shown to describe the target plane
5. Side view showing the normal of the target plane - and the fact that the diagonal isn't correctly aligned prior to op.
6. After op - top face of diagonal now points to target.
To extend diagonal block, use flatten
7. Select bottom face of block
8. Face | Flatten > MMB, select edge shown - or any others parallel to specify axis.
9. Select bottom LH corner vert of RH column to specify length.
10. After flatten op.
To get the desired result result, the diagonal needs to be rotated more about p such that its rear face passes thro' the point r used in (5).
The following stages describe a method of placing a point c in the correct place on the rear face of the diagoal block.
11. Select the bottom rear face on diagonal and mirror to create additional geom.
Use Tools | Connect to draw edges as shown - from vert r to somewhere similar to that shown (not critical) - view in appropriate direction, in ortho when doing this.
12. Select new edges
13. Edge | cut > 10
14. Select all verts along these newly cut edges and apply Vert | Deform > Inflate > RMB. Select vert shown as centre.
![[Image: Ntutdrawcombo5_02.gif]](http://img.photobucket.com/albums/v492/pppix/Ntutdrawcombo5_02.gif)
15. Select vert shown to specify radius.
16. Drag > 100% (hold shift to constrain). You now have an arc of the correct radius (pr - see 10) centred on p. Where this arc intersects the original back face of the diagonal (somewhere between A and B)is the point needed. This pic shows that constructing the arc has modified that back face - and it's no longer a straight edge between A and B.
17. Optional - can cut this edge and repeat the Vert | Inflate op if you want a bit more precision
18. Use Tools | Connect to re-join AB. Doing this provides the point C - the one needed. Can now remove the unwanted geom.
19 - 23. Apply another Rotate to Target > MMB as before using the elements shown.
Still not an absolutely precise w/f - but not too bad imo...and Luiz was asking about 'drawing' type methods
Further pondering ...
No need for 2 rotate to target ops ... just one.
No need for flatten op.
so...
Make 2 columns
Duplicate LH column and extend the bottom - if too long, doesn't matter - not critical.
Create extra geom on rear (as above)
Create arc as above, using point r to set radius.
Once point C has been created, then do the rotate to target op.
RE: Measurements - Need some help - puzzledpaul - 12-11-2012
Another trig based approach - but pythagoras this time.
![[Image: drg.gif]](http://img.photobucket.com/albums/v492/pppix/drg.gif)
Above drawing shows what the wings workflow is based upon.
Black - Left + Right columns.
Blue - Diagonal block... dotted in finish position - I've deliberately made this wider for clarity reasons - the method works for any width of block.
Red arc - centred on p, radius pr
Yellow - radius of arc, intersecting r (and also c).
Triangle xpc (right angled triangle) is the important part of this approach.
Since there are 2 sides whose length is known (bold blue / yellow) - or easily established, it's easy to find the length of the one wanted, xc ... by using Pythagoras' theorem.
The length of XC determines where the dotted red line is positioned [C] from the top of the diagonal. This point [C] ends up at point [r] when the diagonal block is rotated to its final position using a 'Rotate to Target' command.
Pythagoras states - square of the hypotenuse = sum of the squares of the other 2 sides.
Here ...pc^2 = xp^2 + xc^2 ... therefore xc^2 = pc^2 - xp^2
and thus, xc = sqroot [ pc^2 -xp^2] ... which, in 'wings speak' ... in Tab input box, becomes ...
xc = sqrt(pow(pc,2)-pow(xp,2)) ... so, entering the relevant 2 dimensions (pc:bold yellow, xp:bold blue) during the appropriate command will give the required result.
![[Image: Pythagcombo4.gif]](http://img.photobucket.com/albums/v492/pppix/Pythagcombo4.gif)
1. Basic setup - select the 2 verts shown, note distance (pr) is also same as pc. (also measure column width ( not shown). Make the diagonal block a no-segmented object, long enough for the job - but length not critical.
2. Create a mid-way edgeloop.
3. Edge | Slide > Abs ... slide all the way to the block top.
4. Press D to restart slide command (from zero), press Tab, enter relevant numbers into the above expression ... note sign (probably (-) )
Wings will work out the answer to the expression, using the values you've entered ...and locate the eloop in the required position.
5. Select diagonal block, invoke Rotate > Rot to Target > MMB use edge shown for axis > 100%
6. Use same edge or vert on axis for origin.
7. Select target point
8. Specify plane - only 2 of the 3 verts needed are actually shown + either vert behind those shown - note that the thick yellow line in the diagram is a side view of this plane. Similarly, the dotted yellow line shows this plane in its final position after the rotate op.
9. After op.