Dynamic Block Stretch Problems

[ME7CADGuy]

Hi All,

I am trying to make a circular manhole and square/rectangular manhole stretch to whatever size is needed. They both have channels inside the manhole that will need to increase/decrease to match.

The circular manhole channel is just coming up to short or it over shoots its default position within the manhole (I cant find the right number for the distance multiplier for this to work) is there another way to make this work?

The other manhole has more of a complicated channel to make work and I can not think of a way to do this.

We want the channel to stay the same width when the block is increased so using the scale parameter I think will cause issues with this.

Looking forward to any ideas

Blocks attached.

Many thanks

[j.palmeL29YX]

Attached a first attempt for the manhole block. Is this what you want to achieve? 

For the other block, please post an example how it should look after dragging its length/width. Do you have more information about this block (e.g if length/width are restricted to fix values)? 

[Libbya]

To fix your circular manhole block, set BACTIONBARMODE to 0 and then add Stretch2 and Stretch3 to the selection set of the Rotate action.  I'm not sure that there is a simple solution (or any solution) to the square manhole due to the different behavior required depending on where the diagonal channels intersect the edge.  I'd need to think on that further.

[ME7CADGuy]

@j.palmeL29YX  The constraints solution works well for the circular manhole, thank you for this. Is it however possible to achieve without using constraints? I use Autocad LT so if anything needs fixing or amending withing the block I may struggle in future.

@Libbya unfortunately the problem remains with this attempt. If you stretch the manhole to 200 you can see the channel still over shoots the manhole. 

I have shown some examples of what the square manhole should look like after being stretched.

Both these blocks have no size restraints. We want to be able to stretch them to any size that is needed. 

Many thanks

Please see blocks attached.

[Libbya]

If that's not good enough for you, then I think that the best you can hope for for the circular manhole, is to make the center lines a little longer so that they always overlap into the outer ring, then add a many faceted wipeout to that outer ring and adjust the draworder so the center channel lines are behind the wipeout.

Yes, I understood what the square manhole needs to look like in those various states.  That's why I said what I said.  Notice that sometimes the two lines straddle the corner, sometimes they are both to one side and another state (that is not shown, but could exist based on the parameters) would be if the 'height' is stretched so that it is greater than the width so that both diagonal lines end on the vertical sides.  In each of those states, the stretch actions would have need to have different effects on the lengths of those two diagonal lines (sometimes stretching one, sometimes stretching the other, sometimes stretching neither).  Even detecting the transition between those states would be a trick.  You might be able to using some gripless point parameters acted on by the two main stretch actions and then have the point parameters as inputs for a double lookup that switched the visibility state to show the appropriate stretch behavior.  Seems like a bit of a stretch, though, and I don't know that I feel like donating the time required to find out if it is even possible.  I'll probably think on it more, though, as I find this sort of puzzle compelling.  

[j.palmeL29YX]

---

@ME7CADGuy wrote:

If you stretch the manhole to 200 you can see the channel still over shoots the manhole. 

---

What about cheating a little? Fill the "donut" with a solid hatch (color 255,255,255) and hide the unwanted parts of the lines behind the hatch (draworder). 

[Libbya]

---

What about cheating a little? Fill the "donut" with a solid hatch (color 255,255,255) and hide the unwanted parts of the lines behind the hatch (draworder). 

---

I like that idea better than a wipeout.  It brings to mind another option.  Instead of color 255,255,255 you could set screening to 100%, which will erase any linework under it.  That would act the same as a wipeout but with the benefit of allowing true curves.

[j.palmeL29YX · Accepted solution]

A pragmatic proposal for the rectangle manhole: drag the both lines manually rough between the both rectangles. 

Yeah, it is not as comfortabe as you hope to get, but for me it seems to be practicable. 

(and read the tip of @Libbya in post #7 ).

[ME7CADGuy]

Great ideas, thank you both!

I think what you have done to the square manhole @j.palmeL29YX is the best we will be able to get it at the moment.

I added a hatch to the circular one but there is a problem with the hatch due to the scale action, is there a way for the hatch to behave the same way as the boundaries it needs to stay within because the hatch scales bigger than what is needed at the moment? what is hatch screening @Libbya?

I appreciate the effort from you both, great help to the CAD community, thank you.

[j.palmeL29YX]

---

@ME7CADGuy wrote:

 there is a problem with the hatch due to the scale action,

---

You must not add the hatch to the selection set of the scale action. While cadding the hatch be aware that "associative" is ctivated, AutoCAD itself will do the rest.

(Without seeing your current block file I only can guess at the cause...  😞  )

[Libbya]

In the attached file you will find a version of the rectangular manhole that (as far as I can tell) works perfectly and does not use any wipeouts or the like.  It follows exactly what I proposed in message 5.  It took a bit of thinking and effort to get it working properly.  

Regarding screening, I was referring to plot settings and not specifically referring to a hatch setting.  

[j.palmeL29YX]

---

@Libbya wrote:

 a version of the rectangular manhole that ... does not use any wipeouts or the like.   

---

WOW !!

[ME7CADGuy]

@j.palmeL29YXRegarding the hatch for the circular one, I understand what you mean know, completely slipped my mind the hatch will behave like that.

@LibbyaJ is right WOW!! I cant thank you enough for the effort taken to find a solution for this, it is greatly appreciated.

Now to dive a little deeper, do you think it is possible to have a circular manhole with the same channel layout work in the same way or will the scale actions cause problems for this? Do you both think that using a hatch and the solution @j.palmeL29YX had for the square manhole is the best way to go this type of block? or can we go the lookup route?

I am going to try and decipher the latest sqaure manhole soloution, please wish me luck.

I have attched the current soloutions we have plus the new circular manhole with channels for yous to have a little stab if you both dont mind.

Thank you again!

[Libbya]

I know that you said before that you wanted limitless adjustability.  What is a realistic working range?   What sizes would be unreasonably large or unreasonably small for actual use?

[Libbya · Accepted solution]

I went down a different path for the circular ones.  The one that has the single channel does not have any masking. The one that looks like a peace sign has a wipeout at the center.  Hopefully all three now suffice.  

[Libbya]

---

@ME7CADGuy wrote:

I am going to try and decipher the latest sqaure manhole soloution, please wish me luck.

---

It is a bit complicated.  There are three visibility states.  One has the diagonal lines straddling the corner, the second has the lines ending at the vertical edge, the third has the lines ending at the horizontal edge.  The appropriate stretch actions are added to the appropriate diagonal lines so that for their given state they are stretched correctly.  There is a position parameter added to the 10,0 point.  That point is moved left when the vertical linear parameter is made larger, and right when it is made smaller.  That point is also moved left when the horizontal linear parameter is made smaller and right when it is made larger.  That movement means that the transitions between states of the block occur when that point parameter is equal to 6,0 and 11,0.  Between those two points, the lines need to straddle the corner, above 11 the lines both end at the horizontal edge, and below 6 the lines end at the vertical edge.  Turn on the grip for the point parameter if you want to see the behavior in real time.  I then added a double lookup with the appropriate range values to switch the visibility state based on the position of that point parameter.  That would have been enough if the block didn't have the rotate action.  To accommodate that movement and the rotation of the two main linear parameter, I had to use another trick.

There are two gripless chained linear parameters added that are the same length as the main linear parameters and those added linear parameters are scaled by the main linear parameters so that they are always the same size as their respective main linear parameter.  Those chained linear parameter are not rotated and so they have the move actions added to them to move the point parameter.

Hopefully that all makes sense as you look at the block.  If not, then report back with your questions. 

[ME7CADGuy]

Great work from you both, I cant thank you enough for the ideas, support and execution of these. I will implement some of these ideas into our other blocks also.

Truely thank you!