Material Take-Off Precision

[Ilic.Andrej]

Hello,

This question is about precision for creating Material Take-Offs in Revit.

Many joins are being "cleaned up" by Revit. Whenever we use the "Join Geometry" tool, its almost as if we tell it to cut some pieces of geometry or extend other. When we look at the "cleaned up corner"  between two walls, the extent of the wall location line is not always the same compared to the extent of the actual wall which was joined with another. 

This is my question:

How Revit computes material volumes and areas after we have applied the Join Geometry tool and all wrapping materials, slopes etc? I assume that it is not as simple as just using Location Lines, thicknesses and Heights. I would like to have at least some understanding of this to be sure that it is doing it with a desired precision.

[loboarch]

In the simplest terms, the material in the "overlapping" joined geometries will be "discounted" from the material takeoff. But it can be complicated to figure out exactly what is being discounted perhaps.

For example in a "T" connection on  2 walls the wall that is getting "shortened" as it is connecting into the other one at the "T" joint is the one that is reduced, the volume of the crossing wall is not affected. On a simple "T" connection this might be easy to explain or deduce what is happening. In a more complicated joint, it is perhaps not so easy. I can't say I have a summary of the exact coded behavior or logic used. I would say generally if the cleanup/join is altering the geometry as it relates to the position line of the element then that material is getting discounted form  the material takeoff.

Generally I would also say in most cases this is going to fall well within the margin or error for a construction material takeoff.  

[Ilic.Andrej]

Let me understand what you re saying... For area or volume calculation, Revit uses the length of the Location Line as one of the references, but it deducts some quantities depending on the joins? 

[loboarch · Accepted solution]

You are correct. Take these 2 examples below. In one, the wall intersecting the "T" is reduced, in the other the wall is exactly the same length, but not touching the other wall to form a "T" join, so the area/volume is slightly larger.

In a simple connection like that it is easy to understand the behavior/logic applied. In a more complex joint this might not be so easy to see, but the general logic is followed. Thicker walls, miter joints, etc... are going to affect this slightly as well, but again I think the material takeoff is going to be sufficiently accurate to order building materials for a project.

[SJT-72]

I don't accept your argument that this is is accurate enough for construction materials take off. When you have very large buildings with lots of corners the the error gets disproportionately larger. Can you please confirm when this error is going to be fixed?

---

@loboarchwrote:

You are correct. Take these 2 examples below. In one, the wall intersecting the "T" is reduced, in the other the wall is exactly the same length, but not touching the other wall to form a "T" join, so the area/volume is slightly larger.

---

@loboarchwrote:

You are correct. Take these 2 examples below. In one, the wall intersecting the "T" is reduced, in the other the wall is exactly the same length, but not touching the other wall to form a "T" join, so the area/volume is slightly larger.

In a simple connection like that it is easy to understand the behavior/logic applied. In a more complex joint this might not be so easy to see, but the general logic is followed. Thicker walls, miter joints, etc... are going to affect this slightly as well, but again I think the material takeoff is going to be sufficiently accurate to order building materials for a project.

---

In a simple connection like that it is easy to understand the behavior/logic applied. In a more complex joint this might not be so easy to see, but the general logic is followed. Thicker walls, miter joints, etc... are going to affect this slightly as well, but again I think the material takeoff is going to be sufficiently accurate to order building materials for a project.

---

[ToanDN · Accepted solution]

Use Parts if you must have a more accurate quantification.

[loboarch]

I would say the suggestion of parts is the answer if you need more accuracy. Changing this kind of behavior is most likely a fairly significant change to the fundamental way Revit works.

I don't know if I would characterize this as an "error" to be fixed. Could it be better and more accurate, of course, but the way it is working now is the "designed" behavior. 

[SJT-72]

Right, have done a bit more research into this, materials take off for walls gets more accurate if a miter wall junction is used (but this is not always possible). Is there a way to set wall junctions to miter as standard / default, as this would help.

With regards to this not been an error, I have to disagree. Especially when other disciplines are using the Revit model. I do appreciate that it could be difficult to change, but surely the whole point of Revit and a centralised model is accuracy.

Is this a case of its too difficult for the software engineers to rectify, opposed to what the software should be doing by default?

Would welcome other peoples thought on this issue?

[Ilic.Andrej]

So, a wall is like a chunk of geometry, like a simple sweep. It has the location line, thickness, the base and the top constraint. How do you tell a computer to calculate everything that varies among its layers? So like @ToanDN wrote, the walls need to be turn into parts. Makes sense to me. Every layer becomes the independent piece of geometry.

I so pose its possible to program what you want but seams to me that it would take a lot of work. The question is would that cover every possible situation. You can't predict everything. So many different details....

Maybe you can keep the model where the walls are walls and just create a new file from that. Open the copy, turn walls to parts and take the material take-off. After a while and if something needs to be changed, you can open the original model, make the changes and do the whole thing all over again.

[ToanDN]

You don't need to make a copy to create parts. Relationship between parts and original elements is one way, meaning changing parts does not affect the originals, but changing the originals does update parts.

[Ilic.Andrej]

---

@ToanDNwrote:
You don't need to make a copy to create parts. Relationship between parts and original elements is one way, meaning changing parts does not affect the originals, but changing the originals does update parts.

---

I never worked with parts. How can I change the original wall after the parts were made? (I mean stretching, trimming etc...)

[ToanDN · Accepted solution]

The walls are still there unchanged. Set the view to Show Originals, show parts, or show both.

[Ilic.Andrej]

Great!!!

[loboarch]

---

@SJT-72 wrote:

Is this a case of its too difficult for the software engineers to rectify, opposed to what the software should be doing by default?

---

It is a case of wall elements in Revit do not contain "real" geometry. Walls are a somewhat abstracted element. I suspect this was done for performance reason originally, plus it makes dealing with more complicated situations of walls  coming together easier sometimes (but harder in others.)

Flash forward 19 years (Revit was first developed in 1999) with computers today, performance MIGHT (and that is a big might) not be impacted as much if the underlying structure of a wall was changed to be more "real". This is what the parts tool mentioned earlier in the thread attempted to do with wall geometry.  If the wall element itself was able to take on a more part like quality and be more adjustable and controllable the accuracy of the material takeoff of them would be more accurate. Making walls have "real" geometry might have other advantages too?

The issue is the abstract wall element in Revit today is fairly fundamental to the workings of the software, so simply changing it is probably not possible (again i.e. the reason Parts were created). So if a new kind of wall element did not have a performance impact, the development of a "New" wall element could be done, but you probably would have "new" part/layer based walls and "old" abstracted geometry walls. Kind of like when sketch based stairs were replaced by component based stairs. A new element was created that had more functionality. Sketch based stairs code still exists and you can kind of make them, but it happen less and less frequently as the "new" stair tool is used more and more. "New" walls development might looks very similar to that. So it is not as simple as fixing a minor thing with the way walls currently work, it would probably be a fundamental change to or the creation of a totally new element in Revit.

Of course all of this is just conjecture on my part. I am not a developer/coder, I am an Architect by training, but spending the last 13 years in a software company, I have picked up a few things about what is and is not possible or how difficult correcting the "error" in wall material takeoff might be.  😉