Duct Fitting & ASHRAE Table

[CM-Black]

If I create a custom duct tee fitting, how do I associate it with one of the pre-loaded ASHRAE Tables?

[Martin__Schmid]

Interesting to see this old thread have some conversation on it.  The functionality of the XML dates back to the original release, and it wasn't designed to be user extensible... so even if I could articulate  how it works, it may not solve your problem (it's been ages since I've looked at this, so would need to dust away some cobwebs).  

as @iainsavage mentioned, you may be able to accomplish what you need by creating a fitting calculator.. but that doesn't readily solve the problem that you'd likely want to define your fitting families to use that calculator by default.

Can you elaborate more on what problem you're trying to solve, instead of how you're trying to solve it?  E.g., taking a quick look based on the thread above, I see there is no obvious table to use for pressed bends w/ r/D of 1.0 or 2.0... is that the issue, or is it something else?  If I understood what you're trying to do, it may be possible to force a solution (e.g., perhaps adding rows to the XML and/or TBL files) to achieve the desired outcome in another way.

[iainsavage]

As a test I made three small round duct systems, with three different bends: radiused 1D, radiused 1.5D and segmented/gored.

All of them default to CD3-11 which is a fitting that I've never seen in my life.

It says that CD3-1 and CD3-2 are not applicable to the radius fittings but it will let me choose CD3-2 for the segmented/gored bend and conversely it will let me choose CD3-12 for the radius bends. This makes no sense to me at all.

[HVAC-Novice]

the problem I try to resolve is.... i don't like/trust the ASHRAE fitting values and the way Revit selects them. Here my original thread. So I edited elbow families to calculate the pressuredrop based on SMACNA data for fittings. So far so good.

the problem is, Revit by default uses an ASHRAE fitting and calculates that pressure drop plus adds the one I calculate. Obviously i only want to use "my" value and not add the ASHRAE value. To overcome this, I have to manually set the Loss Method to "not defined" and i have to re-do that if I change the type (i.e. change from gored to die-stamped, or from 1D to 1.5D etc.). So what I want is to select that for those fittings no ASHRAE table is used. 

I also came across some fittings (transitions) where I'm OK with the ASHRAE method and don't want to make my own calculation. But there I can choose from multiple fittings and I like to select the one with the highest loss since that comes closer to what SMACNA data show. For those I would like to eliminate the ASHRAE fittings i don't use. Or at least make the ones i want to use the default. 

I'm basically going through my project and manually check each fitting if the calculated pressure drop is reasonable (i.e. close to what SMACNA data suggest). If the oob data are bad (mostly way too low, or really high), i like to add my own calculation. if the result is reasonable, i keep the oob. 

[Martin__Schmid]

As far as how this works...

Table CD3-1 is only applicable for 90deg elbows, r/D = 1.5, from 3" to 10".... you can see that in the table file.. if your fitting meets those constraints, it is a viable selection in Revit, as you can see below.

There is no logic to discern whether a fitting is gored or pressed.. so if the testable conditions (angle, size, r/D) are met, both the Gored and Pressed tables will be presented as options (assuming tables for each exist)... that's why you'll see CD3-2 and CD3-12 interchangeably.

CD3-11 comes up because it's only checking if the fitting is 90 deg, and falls in the range of 3 to 48 inches... this defaults, unfortunately, in many cases, because it catches a lot of conditions, and ends up at the top of the list alphanumerically (CD3-1 and CD3-10 the only options 'before' that don't apply in a lot of cases). 

Thanks for the pointer to the other thread.. a lot there to help dust off the cobwebs.  

Re: "I have to manually set the Loss Method to "not defined" and i have to re-do that if I change the type (i.e. change from gored to die-stamped, or from 1D to 1.5D etc.). So what I want is to select that for those fittings no ASHRAE table is used. " 

To avoid the completely manual intervention, it may be possible to run an automated check of some sort, set this to "not defined" using the API or Dynamo... it may be worth the effort, as it isn't in our near term plans to improve the experience here.. but I do totally understand your comment about trust/confidence.

[iainsavage]

Hi Martin,

If anyone knows the answers its probably yourself.

CD3-2 is a radiused bend of 1D centreline radius but the programme doesn't select it or even allow it to be selected as an alternative option (see the post that I made at just about the same time of your post).

CD3-1 is a radiused bend of 1.5D centreline radius - again Revit won't select that table.

But in my example it allows me to select CD3-2 for a segmented/gored bend.

In all cases Revit defauts to the very wierd (in my experience) CD3-11. 

Here's another example in which the default table is nothing like the actual fitting and pressure drop is 14 Pa:

When I manually change the table to CR3-1 which is more applicable the pressure drop is only 2.4 Pa.

These are just some examples - basically Revit very often seems to select inappropriate tables and the resulting pressure drop calculations therefore tend to be highly unreliable.

Your comment that "The functionality of the XML dates back to the original release" makes sense but there are posts dating back to 2009 on this same topic and its never been corrected.

Basically we should just be able to press the duxt pressure loss report button and have Revit reliably calculate the system pressure drop without having to manually alter settings for fittings, or if that is too much to expect then it would be good to at least be able to understand the logic which causes the programme to select the seemingly wrong values.

[Martin__Schmid]

@iainsavage 

CD3-2 only is an option if size is 3-10", 90deg, and r/D=1.0.... if these are all met, it should be selectable:

See my prior post re: CD3-11... as far as helping you perhaps understand the logic.. recognizing, though, that what you really want is an easier to use solution.

An easy button would surely be ideal... the logic currently in place and lack of user settable defaults is cumbersome to work with, understood.  in order to solve this (including our implementation for the more detailed Fabrication elements) will require an overhaul... and unfortunately, hasn't risen to the top of the list.

[HVAC-Novice]

Thanks for the response. Much of that goes over my head.... and it is sad to hear there are no immediate plans to fix this. I'm finding more fittings that just don't have an ASHRAE reference, so it assumes 0-pressuredrop. This is NOT good. 

My suggestion would be to enable the families to have users override whatever Revit does oob. Or have the family itself select the ASHRAE fitting or include the data in there. That way each family-type always uses the appropriate table or the user value (equation). It kind of makes sense that a fitting that looks "gored" also uses gored values and has higher pressure drop than a die-stamped fitting (that looks smooth). Same for a 1D vs. 1.5D. If for some reason a 1.5D doesn't fit, the higher pressuredrop of the 1D should show up. 

I'm trying to optimize the duct system to use better fittings and sizes so my pressure is lower and branches are somewhat balanced (inc. static regain). This obviously doesn't work if the fittings are wrong. oob it shows me that using 2 90° fittings has much lower pressure drop than using two 45° and so on. The way it is, the critical path could be the non-critical path, or not. I don't expect 100% accurate values, but they should at last make sense.

[iainsavage]

Martin regarding "Table CD3-1 is only applicable for 90deg elbows, r/D = 1.5, from 3" to 10".... you can see that in the table file.. if your fitting meets those constraints, it is a viable selection in Revit" I'm not getting this behaviour as shown below even although all criteria are met. Does this only work with imperial units/families?

[Martin__Schmid]

@iainsavage - see metric example below... using a different family than what was in your screenshot, yes, it works as expected. 

My suspicion is that the way that slip joint elbow from your example is defined throws things off (see 2nd image below w/ explanation).

What is happening is that when determining r/D.. Revit is looking at the connector locations.. not specific parameters.

Since there is the extension for the slip joint in that family... it changes where the connectors are, which thus affects the determination of r, which affects r/D... e.g., the r ends up as shown by the red solid line.. which you can see is longer than the labeled Centre Radius dimension... thus, making r/D larger than 1.5.

Testing that hypothesis.. if you change the formula in that family as shown below... and reload into the project.. the CR3-1 option becomes available.

Thanks for raising this issue.. I don't think I've seen/heard this scenario before.. more good fodder for considerations for potential overhauling how this works.

[iainsavage]

Martin,

It all makes sense when its explained by someone who knows the innards of the software, but just to let you know that fitting is the Revit out of the box fitting in the UK metric library.

I’ve had trouble with this for a long time but now I know how to “fix” it.

Many thanks for this.

PS: if I delete or rem out sections of the ashrae.xml file would that get rid of some of the stranger fitting options which I never use, such as CD3-11?

[HVAC-Novice]

FWIW, today it seems some of the ASHRAE fittings disappeared after i yesterday removed their XML-entries AND the file for each of the fittings. Maybe Revit just needed to restart some more times to update this.

I will try more, but so far it looks like unwanted fittings can be removed. I don't know if the XML editing alone would help, but i just do both. Have to do that again when Revit updates to 2023. 

I think if we can reliably remove ASHRAE-fittings, I may be able to use "my" fittings with less hassle. 

edit: you actually can simplify the equation in the family if you set the fitting to use the coefficient. if you look in my family, you see I made a detour to calculate velocity and velocity pressure. I ran a pressure report and it shows a pressuredrop and the coefficient that I set. So this PROBABLY works fine. It prevents me from seeing the pressuredrop while in the model, i only see it in the report. 

I tied the flow value to a parameter so I can see in the model if it gets flow, but in the properties it shows 0 (not sure why since it clearly calculates a pressuredrop per the report). One issue is, i can't tie the calculated pressure to a shared parameter that i can show in a tag. 

[iainsavage]

I’m also wondering if I can tweak the formulas in the xml file to introduce a tolerance band for r/D ratio etc, maybe 0.9 to 1.1 or something - seems like at the present time if its not an exact match then it gets passed by in the xml logic.

I’ll have a play about later - lunch break over, better do some actual paid work now!

[Martin__Schmid]

Sure thing, @iainsavage -- yes, I noticed that file was a shipping piece of content.  

Interesting question... I got rid of the CD3-11 related row, and noticed no change in behavior.  I then got rid of ASHRAE.XML file altogether, and the functionality similarly doesn't seem to change. If I remove the CR3-11.tbl file (e.g., CR3-11.tbl), I end up getting very unexpected results when that option is selected (-5000 in wg).

If I had to guess, I'm assuming that the contents of the XML file are actually encoded into the Revit code (perhaps for performance reasons)... I also tried w/ an empty template file to see if it was perhaps cached in there... but same results as above...  thus, seems editing it will have no effect.. at least not in my poking.. Perhaps the file itself is just there as part of the data provided with all the other ASHRAE Duct Fitting Database.

[iainsavage]

The large print giveth and the small print taketh away (Tom Waits)

[iainsavage]

Jumping back to part 29 above, I modified the OOTB bend and it now works.

This is the OOTB fitting:

This is the modified fitting with the connectors moved onto the ends of the radiused sweep:

I now get CD3-2 as an option:

This is also dimensionally more accurate because in reality the duct would extend over the slip joint rather than butting up to it.

With this one from part 25 above:

I suspect that as the duct turns from vertical to horizontal Revit is interpreting it as swapping the width and height dimensions, although this doesn't seem to be the case in my model - both ducts have 800mm width x 400mm depth but I think Revit sees one of them as being 400x800.

Thanks again @Martin__Schmid for giving me some insight into why these things might not be working as expected.

[Martin__Schmid]

@iainsavage re: message 25 about SR3-1.. unfortunately, again, this is a case of the order in which the table checks are being made, vs. having a way to explicitly pre-define which table to use as part of the content definition... computationally, it is doing everything 'correct'... with the limitation that it has no way of knowing that the corner is smooth vs. mitered in order to exclude the mitered table.

The inversion of the WxH is correct in this case as demonstrated below to ensure the values 'map' to the orientation of the reference fitting.

[iainsavage]

So basically, Revit doesn't automatically assign the correct table and we always need to manually select the fittings and assign the correct one from the list of possible options.

For a raduised rectangular bend (OOTB) it defaults to this (unless the angle is less than 90 degrees):

When it should actually be this:

Pressure drop of the default fitting is nearly 5 times as much as the correct fitting.

and you can't make this change en-masse because if you select more than one fitting it doesn't present any choices at all:

I wonder how many systems have been wrongly sized because of this.

@Martin__Schmid maybe you could persuade someone at Autodesk to overhaul this and come up with a better methodology?