Trying to understand the constraints in this block

[h_s_walker]

Ok so this is the 'I' beam block from the standard drawings. 

If you change the radius you will see that the centre of the arc moves outwards so the arc "stays" where it is.

Now every time I've tried to do that the centre of the arc stays where it is so the arc "moves"

I don't want this as I'm trying to create a similar block, but with a sloping flange.

[Libbya]

The widths of the flanges and the middle web are controlled using dimensional constraints and geometric constraints to make the two flanges equal to each other.  There is no way for the arcs to do otherwise.

You likely already know this, but the exact same functionality can be done without using any constraints.  Let me know if you have any questions about that or want to see an example.

[h_s_walker]

@Libbya Yeah I understand the basics. The problem I have is if the radius changes size the midpoint moves in the I-beam block, but when I try it the midpoint doesn't move. See the block to the right in the attached drawing. The block I'm going to try and create eventually is the tapered flange beam, but I can't get started if I can't get the midpoint to move properly

[j.palmeL29YX]

---

@h_s_walker wrote:

See the block to the right in the attached drawing. 

---

Do you mean the left sketch in the attached image? 

If so, post your dwg what you have until now and we will see what is going wrong. 

[j.palmeL29YX · Accepted solution]

Attached you find an example how I would build such a block. 

[h_s_walker]

@j.palmeL29YX That's almost perfect. I could never get the centre of the arc to move so that the fillets stayed where they were.

The almost perfect is I just need to rotate the diagonal lines a little bit so they're a five degree slope.

Thanks

[j.palmeL29YX]

---

@h_s_walker wrote:

... I just need to rotate the diagonal lines a little bit so they're a five degree slope.

---

The choosen dimensions and also the values in the table are for this example of course only fictive. Depending on the beam type and the standard (I've choosen any DIN Standard) you must adjust all these value to your needs.  😉 

Perhaps you could still post your attempt so we can see what you did wrong (and avoid this mistake in the future) - if you're interested. 

[h_s_walker]

@j.palmeL29YX I just changed the dim angle to 5 to get what I wanted.

Ooh I see in my previous post I forgot to attach my start attempt where the centre of the arc didn't move. I've attached the drawing here

[j.palmeL29YX]

Your sample geometry has too few constraints (or too much freely movable geometry). In the current case, AutoCAD has several options for responding to a new radius. In my experience, AutoCAD usually seems to choose the option that allows the geometry to be changed with the least (internal) effort.
If you add further constraints (e.g., fixing the position of the lines), you will get the expected result.

Therefore, you should always aim for fully defined geometry. (You can check it with BCONSTATUSMODE = 1)

[h_s_walker]

@j.palmeL29YX I was using the built in Autoconstrain button. The attached is what I started with and as you can see the arc moved the wrong way

[h_s_walker]

@j.palmeL29YX Is there a way to set the minimum flange thickness to stop the image below happening Something like Original T+(current B / original B)?

[j.palmeL29YX]

"I was using the built in Autoconstrain button"

I would not recommend applying AUTOCONSTRAIN to complex geometry. Too often, illogical constraints are created that do not correspond to my intentions. Or unwanted constraints are created simply because, for example, lines have been drawn almost parallel or perpendicular, but are not supposed to be parallel or perpendicular at all. Finding and correcting such errors is often more tedious than manually adding the desired dependencies step by step. And to come back to your example: as long as the geometry is not completely defined, it is always possible that it will not behave as intended. AutoCAD itselfe decides how to solve the request (the changed radius). But after you have constrained the position of the both lines, changing the radius will show what you expect.

[j.palmeL29YX]

---

@h_s_walker wrote:

Is there a way to set the minimum flange thickness to stop the image below happening Something like Original T+(current B / original B)?

---

Post your block please. In the image I can not see how you drive the flange's thickness. 

[h_s_walker]

@j.palmeL29YX It's your block. If you make the width of the flange 250 you will see that the vertical lines go below the bottom of the flange. See the image below

[j.palmeL29YX]

Hmm, >here< (FYI) you find the table where I grabbed the values from. The greater the width (b), the greater also must be the value for tg. In the attached dwg I added a line for the largest (standardized) size. If you want to build a block whose width reaches to 250, you of course also must increase the value for tg and/or decrease the value of the angle. 

BTW: You can restrict the min/max values for dimensional constraint parameters, but that's not the issue here. 

[h_s_walker]

It's not the table that's the problem.

I'll be changing the block to use the grips and the properties manager.

The problem is if you change the width of the beam the flange thickness stays the same and eventually instead of the vertical edge of the flange being above the bottom of the flange it stretches to below the flange and vice versa for the top flange

[j.palmeL29YX]

What should happen if you exceed the maximum "allowed" value when dragging the width?

[h_s_walker]

It's not a maximum allowed. If you look at the image below. If you make the width of the flange wider than a certain distance the diagonal part of the flange goes UNDER the bottom of the flange,

[j.palmeL29YX]

---

@h_s_walker wrote:

It's not a maximum allowed.

---

[h_s_walker]

When dragging if that point is reached I would like the flange to automatically thicken. so that the flange will always look like a flange