Definition of Draft Angle in Help not correct

Rhino 4.0.

Help says “The draft angle depends on the construction plane orientation. When the surface is vertical/perpendicular to the construction plane, the draft angle is zero. When the surface is parallel to the construction plane, the draft angle is 90 degrees.”

This is not what Rhino actually calculates. What Rhino actually calculates is as follows:

Let Dir be the Direction of the surface. Let N be the normal to the C-plane.

If D is the same direction as N, then the Draft Angle is 90 degrees.
If D is perpendicular to N, then the Draft Angle is 0 degrees.
If D is the opposite direction to N, then the Draft Angle is MINUS 90 degrees.

Do you need to be convinced of this?

Make a hemisphere, oriented like a bowl. Copy it, and rotate the copy 180 degrees, so that it is oriented like a dome. Do Draft Angle Analysis, and set the range to -90 degrees to 90 degrees. The bottom of the bowl and the peak of the dome show opposite ends of the blue-to-red scale, even though the surfaces at those two points are both “parallel to the construction plane.”

Please fix this. I wasted time being mystified by what I saw and then figuring out what happened.

Hi David - what does the Dir command show you as a surface direction (white arrow) in each case?

-Pascal

The Dir arrows are toward the center of the object each time. (Down at
the apex of the dome, up at the bottom of the bowl.)

(The bowl is the result of actually rotating the dome 180 degrees around a
horizontal diameter.)

In a message dated 4/7/2016 18:02:58 Eastern Daylight Time,
steve@mcneel.com writes:

     _pascal_ (http://discourse.mcneel.com/users/pascal)  McNeel  

April 7
Hi David - what does the Dir command show you as a surface direction
(white arrow) in each case?
-Pascal

Visit Topic
(Definition of Draft Angle in Help not correct) or reply to this email to respond

To stop receiving notifications for this particular topic, click here
(http://discourse.mcneel.com/t/definition-of-draft-angle-in-help-not-correct/31
242/unsubscribe) . To unsubscribe from these emails, change your user
preferences
(http://discourse.mcneel.com/my/preferences)

Yeah - if you Flip one of the objects and then ClearAllMeshes, they will agree, since the read of the pull direction also takes into account surface direction - if you instead of flipping, OffsetSrf > Solid=Yes then you’ll see the up facing surface of each (+z) has a different display than the bottom (-z) facing one but the ups match each other…

-Pascal
.

Uuuuuuh, Pascal:

Read my original message.

The complaint is that the behavior of Rhino does not match the
documentation.

One or the other should be fixed.

David Golber

In a message dated 4/7/2016 23:46:04 Eastern Daylight Time,
steve@mcneel.com writes:

     _pascal_ (http://discourse.mcneel.com/users/pascal)  McNeel  

April 8

davidgolber:

The Dir arrows are toward the center of the object each time. (Down at the
apex of the dome, up at the bottom of the bowl.)

Yeah - if you Flip one of the objects and then ClearAllMeshes, they will
agree, since the read of the pull direction also takes into account surface
direction - if you instead of flipping, OffsetSrf > Solid=Yes then you’ll
see the up facing surface of each (+z) has a different display than the
bottom (-z) facing one but the ups match each other…
-Pascal
.

Visit Topic
(Definition of Draft Angle in Help not correct) or reply to this email to respond

In Reply To
davidgolber (http://discourse.mcneel.com/users/davidgolber)
April 8 The Dir arrows are toward the center of the object each time.
(Down at the apex of the dome, up at the bottom of the bowl.) (The bowl is
the result of actually rotating the dome 180 degrees around a horizontal
diameter.) In a message dated 4/7/2016 18:02:58 Eastern Daylight Time,
steve@mcneel…

Visit Topic
(Definition of Draft Angle in Help not correct) or reply to this email to respond

To stop receiving notifications for this particular topic, click here
(http://discourse.mcneel.com/t/definition-of-draft-angle-in-help-not-correct/31
242/unsubscribe) . To unsubscribe from these emails, change your user
preferences
(http://discourse.mcneel.com/my/preferences)

You are correct. The documentation is wrong (or at least it is not very clear)

The documentation is either assuming that only surfaces with normals pointing up will be analyzed or that negative draft angles don’t exist. Either way that’s not correct.

The display itself has to differentiate between positive and negative draft to be useful. If the display regarded plus 2 degrees draft as equal to minus 2 degrees draft that would make the display completely useless for analyzing draft angles.

@margaret, could you get this updated for current help files?

Perhaps Rhino should take the max and min draft angle from the object(s) and initially populate the box with these values. Scan and Solve FEA does something like that for max and min stress / deformation / … The user can then always input values that appear appropriate to that user in that situation.

The draft angle analysis tool could do with some improvement. I had a demo version of Rhino Mold for a while and found its version much better; unfortunately I couldn’t justify the cost of the plug-in so have gone back to relying on the native Rhino tool.

I tend to use it just to check the split line of a component, not the specific draft on individual faces. I set the range to 0 and it gives a reasonably clear indication of where the split is. It’s handy for spotting faces with draft going in the wrong direction. It’s limited but useable.

Hi David. I understand. It seemed from your comments that you were mystified by the behavior, is all - I was attempting to point out what is going on.

-Pascal

Pascal, can you make a YT item for this. I am not clear what to put into the help that would improve things.

“I am not clear what to put into the help that would improve things.”

You could state what the thing actually calculates. Quoting myself:

Let Dir be the Direction of the surface. Let N be the normal to the
C-plane.
If D is the same direction as N, then the Draft Angle is 90 degrees.
If D is perpendicular to N, then the Draft Angle is 0 degrees.
If D is the opposite direction to N, then the Draft Angle is MINUS 90
degrees.
(Of course, I am assuming - perhaps incorrectly - that “the normal to the
C-plane” is clearly defined somewhere else … after all, an un-oriented
plane has two unit normals.)
If you can find some sweeter way to say this go ahead. Just make sure
that what you say is 100% true. What’s in the Help now is not.
David Golber

In a message dated 4/8/2016 13:20:28 Eastern Daylight Time,
steve@mcneel.com writes:

     _margaret_ (http://discourse.mcneel.com/users/margaret)  McNeel  

April 8
Pascal, can you make a YT item for this. I am not clear what to put into
the help that would improve things.

Visit Topic
(Definition of Draft Angle in Help not correct) or reply to this email to respond

In Reply To
pascal (http://discourse.mcneel.com/users/pascal) McNeel
April 8 Hi David. I understand. It seemed from your comments that you
were mystified by the behavior, is all - I was attempting to point out what
is going on. -Pascal

Visit Topic
(Definition of Draft Angle in Help not correct) or reply to this email to respond

To stop receiving notifications for this particular topic, click here
(http://discourse.mcneel.com/t/definition-of-draft-angle-in-help-not-correct/31
242/unsubscribe) . To unsubscribe from these emails, change your user
preferences
(http://discourse.mcneel.com/my/preferences)

That would work,

A simpler statement would be that where the surface normal is pointing up the draft angle is positive and when pointing down the draft angle is negative and when horizontal the draft angle is zero. Up and down are defined by z=axis of the current cplane when the draft angle analysis command is started.

2 Likes

@margaret, @davidgolber - I am not sure - Help has this:


The pull direction for DraftAngleAnalysis is the z-axis of the construction plane that is in the active viewport when the command starts.

The normal direction of the surface is the same as the pull direction of the mold. You can check this with the Dir command.

Changing the construction plane before using DraftAngleAnalysis lets you define any direction as the pull direction.


We could re-word that perhaps:

The normal direction of a the object ( Dir command, Analyze menu > Direction) should correspond to the side of the mold that will be pulled from it.

??

-Pascal

There are those of us that are using this that are not making molds, and
can only guess at what “pull direction” means. I have a vague idea of what
"draft angle" means, but I was using DraftAngleAnalysis to understand the
location of the steepest point of a surface that I am working with, and to
see just how steep it was.

You should define what a Rhino operation does using Rhino concepts, without
assuming that the user is a mold maker, a brick maker, a lawyer, or …

Personally, I haven’t done much with construction planes … I just use
what’s there in the usual set-up.

If you feel that “the z-axis of the construction plane” is well-defined,
then you can say

Let D be the Direction of a surface at some point. Then the "draft angle"
of the surface at that point is
90 degrees if D points in the same direction as the z-axis of the
construction plane,
0 degrees if D is perpendicular to the z-axis of the construction plane
-90 degrees if D points in the opposite direction from the z-axis of the
construction plane,

As for the suggestion:

"The normal direction of the surface is the same as the pull direction of
the mold. You can check this with the Dir command."
I can only ask "What mold? You mean that green stuff on my piece of
cheese?"
What is the noun to which “this” refers in your suggestion? I can’t
tell.
Actually, I can guess what you are trying to say … and you are far, far
from being precise.
(1) A mold is a three-dimensional solid object, usually made out of
metal. There are no pieces of metal in Rhino.
(2) A surface can be used to define TWO solids, one on this side of the
surface, and one on that side of the surface. And in fact, I am using
the same surface in both ways: One to define the object I am making, and one
to define the cradle in which to hold it while I am working on it.
(3) A surface has TWO (unit) normals, pointing in opposite directions.
So I suggest you stop trying to talk about pieces of metal, and just
carefully tell the reader what Rhino does, in terms involving only Rhino
objects… Remember: There are no molds in Rhino.
Dave Golber

In a message dated 4/8/2016 17:31:24 Eastern Daylight Time,
steve@mcneel.com writes:

     _pascal_ (http://discourse.mcneel.com/users/pascal)  McNeel  

April 8
@margaret (http://discourse.mcneel.com/users/margaret) , @davidgolber
(http://discourse.mcneel.com/users/davidgolber) - I am not sure - Help has
this:
The pull direction for DraftAngleAnalysis is the z-axis of the
construction plane that is in the active viewport when the command starts.
The normal direction of the surface is the same as the pull direction of
the mold. You can check this with the Dir command.
Changing the construction plane before using DraftAngleAnalysis lets you
define any direction as the pull direction.
We could re-word that perhaps:
The normal direction of a the object ( Dir command, Analyze menu >
Direction) should correspond to the side of the mold that will be pulled from it.
??
-Pascal

Visit Topic
(Definition of Draft Angle in Help not correct) or reply to this email to respond

In Reply To
margaret (http://discourse.mcneel.com/users/margaret) McNeel
April 8 Pascal, can you make a YT item for this. I am not clear what to
put into the help that would improve things.

Visit Topic
(Definition of Draft Angle in Help not correct) or reply to this email to respond

To stop receiving notifications for this particular topic, click here
(http://discourse.mcneel.com/t/definition-of-draft-angle-in-help-not-correct/31
242/unsubscribe) . To unsubscribe from these emails, change your user
preferences
(http://discourse.mcneel.com/my/preferences)

Well. Keep in mind it’s not a text book, it’s a help topic. It seems reasonable, in the context, to assume users know what a mold is, and what draft angle is. Let’s accept that that is the case. Just for now. Does my proposed rephrasing then seem useful?

-Pascal

In a message dated 4/8/2016 18:53:06 Eastern Daylight Time,
steve@mcneel.com writes:

Well. Keep in mind it’s not a text book, it’s a help topic. It seems
reasonable, in the context, to assume users know what a mold is, and what draft
angle is. Let’s accept that that is the case. Just for now. Does my
proposed rephrasing then seem useful?
-Pascal

It’s in the Help of Rhino. It documents a Rhino function, just as the
same Help documents Surface Two Rails or any other Rhino facility. Yes, it’s
NOT in a textbook about mold making. I used it, and I was not making a
mold. The description absolutely should not reference molds.

No, your proposed rephrasing is NOT useful.

Let me point out: you just wrote “in the context” … What context do you
mean? I think you are referring to your special knowledge that this
function is used by mold makers. But that was not my context. I used it and I
am not a mold maker.

Would you write in the documentation: “This function is designed to be used
by mold makers, and the description is intended to be comprehensible only
by mold makers”? Of course not … but that’s what you mean!

You work for Rhino. You write software. You don’t make molds, cut wood
(what I do), mix chemicals. Please do your job well, and don’t assume you
know so much about other people’s work.

Really: What’s wrong with my description:

Let D be the Direction of a surface at some point. Then the "draft angle"
of the surface at that point is
90 degrees if D points in the same direction as the z-axis of the
construction plane,
0 degrees if D is perpendicular to the z-axis of the construction plane
-90 degrees if D points in the opposite direction from the z-axis of the
construction plane,

It’s precise, accurate, and entirely in the Rhino domain. Is it too
wordy? Yes, it has more words than the current Help. But it has the advantage
of being precise and correct.

In conclusion (which means I’m getting tired of this conversation, and wish
you worked for me) let me quote Einstein: “A thing should be described as
simply as possible … but no simpler.”

Let me remind you: This began with me complaining that I spent a extra
half an hour figuring out how to use this function. My time was wasted
because of the bad description. I’m the customer. Your job is to enable me to
do my job without unnecessarily wasting my time.

David Golber

Oh my. Bad day? I’ll see if we can clarify the wording in the help topic so that it is readable and clear enough.

thanks,

-Pascal

1 Like

That statement is not very useful or accurate.

David is correct in stating that you are looking at this in a completely wrong-headed way. I am a pattern maker and pattern makers have been using draft angles long before any injection mold builders ever existed. so the idea that draft angles and draft angle analysis was invented for injection mold builders is bogus.

That said the people like myself that use draft angle analysis every day are not even going to read the help files so don’t try to imagine what they want to read since they won’t be reading it anyway.

David is correct in stating that the problem with your approach to an explanation is that it assumes a mold component exists and that said mold has all its relevant surfaces configured so that they are facing one direction. If the mold maker already knows those facts then its not likely he will even need to be analyzing draft angles. Mold makers don’t just use draft angle analysis to verify that an simple mold design is functional and to suggest that is the commands main purpose is not accurate. Of coursem as I said, the moldmaker is not going to read the Help so he won’t be swayed by what the help file says he “should” be doing.

The point of draft angle analysis is that you can apply it to any object which has normals pointing in whatever direction (if one already knows all the directions there would be no need to analyze). You can analyze the draft angles in whatever range of angles that you might be interested in, relative to the current cplane Z axis. And that is information that the help file should convey.

@davidgolber

I bet your chances of getting a positive answer (and helping to actually improve the help file) would have increased greatly if you had not written that last message and made yourself sound rather… unpleasant. But then again… @pascal is such a nice, polite and helpful man, that he will probably see to it anyway. But please keep your “I’m the customer and I’m always right - you should do my bidding”-rhetorics to yourself. This forum is based on keeping a polite tone and respecting the massive amount of work that both McNeel employees and users around the world do, so that we all can get work done, learn new things and help improve Rhino.

My 2 cents.

-Jakob

2 Likes

Three suggestions:

  1. Add a line to the help file for DraftAngleAnalysis which states that draft angles can be negative as well as positive, and the sign depends on the direction of the surface normal.

It may not be obvious to a user that a draft angle can be negative.’ If the color display range is set with 0 as the minimum any areas with negative draft angle will appear to have zero draft angle. This leads to the second suggestion:

  1. Add options to the DraftAngleAnalysis display panel for “Auto Range” and “Max Range”, the same as currently used for CurvatureAnalysis.

  2. Delete from the DraftAngleAnalysis help text the sentence that “Draft angle is used to design injection-molded parts that must eject from molds.” As others have stated above it is also used for many other purposes.

2 Likes