Turning Vanes and Duct Elbows, Part 2

In the previous post, we were looking at how the number of turning vanes in an elbow impacted its performance and discovered that there are a lot of variables to consider.  For those who have never actually seen turning vanes, here is a picture of an elbow before it was installed on a recent project.

Here is a close up of this particular turning vane design.

As you can see, this particular vane is a double wall vane that is mounited to the duct via a rail assembly. (The link takes you to a specification sheet for this product that will show you more detail if you are interested).

The fact is that the specifics of the turning vane design can make a big difference in the loss through an elbow.  ASHRAE documents about 11 different turning vane designs in their data base and there is a factor of 4 difference between the lowest pressure drop and highest pressure drop design.  Surprisingly, double wall vane designs tend to have higher losses associated with them when compared to some of the single thickness designs.

Returning to our previous discussion, one variable that we didn’t change (other than to experiment with aspect ratio) was the size of the duct. In other words, we looked at a duct with a relatively low flow capacity. The following table illustrates what happens if you take the same fitting shape and scale it up to handle a lot more air.

As you can see, a poor fitting that was of little consequence in a small duct could have a
significant loss in the large duct. This is because small ducts take a lot more perimeter to surround the volume they contain.

Since most of the friction associated with moving air through a straight duct is related to the interaction of the air with the duct wall at the perimeter, then a small duct must operate at a lower velocity than a larger duct of the same aspect ratio if the friction rate is to be held constant.  As a result the change in carrying capacity associated with changing a
duct size is not directly proportional to the change in duct cross section.

Lower velocities imply lower velocity pressures, so the impact of a poor fitting in a small duct will be much less significant than if that same fitting were installed in a larger duct operating at the same friction rate.

Now, consider for a minute how duct designs are conveyed to the field in our industry for the purpose of bidding work. The drawings are always “to scale”. So, on a two line duct drawing that illustrated a 12″ x 12″ three vane version of the elbow geometry we
have been discussing, the elbow would be 1/2″ wide when drawn on a 1/2″ to the foot scale plan detailing a mechanical room.

If that same elbow were drawn on 1/8″ to the foot scale floor plan showing the distribution system, it would only be 1/8″ wide. At that size, it would be very difficult to convey the details of the vane geometry that are critical to controlling the losses through it. If the drawings were one line drawings, it would be impossible. In fact, on a one line drawing, both the 60″ elbow and the 12″ elbow illustrated in the table would look the same.

Without further guidance, a tradesperson fabricating the duct could easily decide to build both elbows with out turning vanes, especially if they are under pressure to complete the job quickly and bring it in under the budget established by the competitive bid that won their company the work. But, the large elbow with no vanes could impose an unacceptable loss on the system it served if the designer had anticipated that it would be fabricated with them.

This could be a really big deal if there were a number of such elbows. In contrast, the impact of the smaller elbow fabricated with out vanes may not be significant enough to show up as a performance problem even if the designer had intended that vanes be provided in all locations.

The bottom line is that the losses through a fitting are very dependent on a wide array of variables, many of which will impact its cost. All of this can seem a quite overwhelming at first, especially if you are under time and budget pressures; how can you analyze each elbow and make the best decision?

The fact is that you probably can’t. But, by asking questions like the one that started this string of posts and gaining an insight into the critical factors that are associated with the losses through a fitting, you will gain the ability to make good judgment calls with out the analysis.

Plus, with computers and software, there are some neat tools to help you make a very informed decision in a timely fashion. The numbers I used for most of the tables in the preceding post were generated using the ASHRAE duct fitting database software in a matter of a few key strokes; I spent more time writing about and printing the results than I did doing the math.

In my opinion, a tool like the ASHRAE duct fitting data base is well worth the
investment because it allows you or your design team to make very good decisions with very little effort. And as you and your team use it, your “gut” learns and soon, you can make good decisions on the fly with out the software. This is a skill that is very handy
out in the field if you are trying to decide if a fitting needs to be re-done on a construction walk-through or a retro-commissioning/troubleshooting project.

Clicking here will link you to the point on the ASHRAE website where you can find out more about the tool. The ASHRAE handbooks have similar information in tabular form; very useful but a bit more time consuming.

United McGill publishes loss coefficients for their products and you can download their literature from their web site.   If you register with them for their electronic newsletter, I believe you will receive a free copy of their duct design CD, which is very good and probably worth what they would charge for it even if you couldn’t get it for free.

As you may have gathered from this discussion, duct fittings need to be considered in the context of the system they serve. AMCA International “publishes and distributes standards, references, and application manuals for specifiers, engineers, and others with an interest in air systems to use in the selection, evaluation, and troubleshooting of air system components” (the quote is from their web site; the emphasis is mine).

For example, AMCA publishes data on the impact of fittings on fan performance when the fittings are installed in the immediate vicinity of the fan inlet or outlet. I have found their publications 200 – 203 to be particularly valuable and well worth the investment. Their biannual publication “In Motion” can also be downloaded from the site for no cost and contains informative articles and a listing of AMCA members.

David Sellers
Senior Engineer – Facility Dynamics Engineering

Click here for an index to previous posts

This entry was posted in Uncategorized. Bookmark the permalink.

6 Responses to Turning Vanes and Duct Elbows, Part 2

  1. Bruce Savage says:

    David, thanks for the article Turning Vanes and Duct Elbows (2011)

    Your article caught my attention because of the use of a single vane in your low speed tests.

    I have a slightly different problem that I was hoping you might provide some insight. In my house there is a branch duct sized at 8″x20″, The duct has a 90 degree horizontal elbow turning on the 8″ side. The outside corner is curved but the inside corner is not. From what I have read it appears a sharp 90 on the inside is the major cause of turbulence and the curved outside edge is of little benefit to address this. It also appears the amount of the turbulence created at the bend is significantly impacted by the width of the duct (20″) not its height. On sites that provide an equivalent “additional pipe length” the impact of a sharp 90 at this width is shown to be as much as 5 times that with a curve inside corner.

    Unfortunately installation of vanes at this location is limited to what I can reach and at best that would be about the first 5″ of the elbows inside radius. I thought the impact of the corner might be reduced by splitting the elbow into two sections (5″ and 15″) with an extended vane. This would create a 5″ sharp elbow and a 15″ full radius elbow. Do you know if this would be of benefit and if so, is there an optimal position for the single vane?

    I appreciate your thoughts on this.


    • Hi Bruce,

      I’m glad you found the post to be useful. Sounds like you are doing science experiments with your house, which I do a lot and learn a lot from.

      One thing I wanted to be clear on is that the data I presented is based on a calculation using loss coefficients from the ASHREA Duct Fitting Database, not test data. Because of the turbulence coming off of the leaving side of the fitting, it would be difficult if not impossible to measure the loss for a given fitting in the field and challenging in a lab setting.

      That said, The ASHRAE Duct Fitting Database losses tend to be supported by the pressure gradient measurements I have made, meaning the pressure gradient I predict with calculations from the database correlates pretty well with field measurements of the actual pressure gradient if I correct the field data for the flow rate used in the calculation (or vice versa). So, I have a fairly high level of confidence in what they predict.

      The losses through any fitting aver very dependent on the velocity (which is a function of flow rate) through it. Meaning that a poor duct fitting that is operating at a very low velocity may not really represent much of a loss. But that same fitting, operating at a high velocity, could be a disaster. That is primarily because the fitting losses are a function of velocity pressure and the velocity pressure varies as the square of the flow; double the flow rate and the velocity pressure will go up by a factor of 4. That relationship falls out of the Darcy Weisbach equation, which is kind of cool to say – in a nerdy sort of way – if you are discussing this. That is what I am trying to illustrate in the second table presented in the blog post leading up to the one you commented on. Note how doubling the flow (and thus the velocity) turns a relatively inconsequential loss into a significant loss.

      So, to really answer your question, I would need to have a sense of the flow that is supposed to be going through your elbow. That said, ASHRAE has the fitting you describe with a single turning vane in it included in the database. What they don’t have is the fitting with out a turning vane. But to get a sense of what your improvement might be worth in terms of a loss, I hypothesized that your fitting with out a turning vane would probably approach the losses you would get from a mitered elbow with out turning vanes; probably a bit better than that, but at least it gave me a frame of reference to use.

      So, when I ran the numbers in the fitting database, here is what I came up with. The loss coefficients pretty much tell the story and interestingly enough, the mitered elbow is about the same as the fitting you are dealing with. The smooth radiused elbow is nearly an order of magnitued better

      The Mitered Elbow, No Vanes is pretty much what it says. The Square Throat Radiused elbow is the fitting you are describing with one turning vane. The Radiused Elbow with No Vanes is also pretty much what it says and is one of the elbows I was using in the example in the blog post.

      Flow, cfm Velocity, fpm Mitered Elbow, No Vanes Square Throat Radius Radiused, No Vanes
      Loss Coefficient 1.03 1.00 0.14
      555 500 0.02 0.02 0.00
      1,111 1,000 0.06 0.06 0.01
      1,667 1,500 0.14 0.14 0.02
      2,222 2,000 0.26 0.25 0.03
      2,778 2,500 0.40 0.39 0.05
      3,333 3,000 0.57 0.56 0.08

      As you can see, at low velocities, none of the elbows are that different in terms of pressure drop. At higher velocities, they start to diverge.

      In my experience, most residential duct systems operate at velocities in the 500 -1,000 fpm range, maybe 1,500 fpm at the most. If that is true for your system, then modification you propose would provide little measurable benefit, especially given the number of curse words it would take to do it.

      If your velocity was at the high end of the scale I associate with residential systems (1,500 fpm), an elbow with a radius at the heal and far side would start to offer some benefit. But the centerline radius would need to be in the range of 12 inches, which probably means some sort of offset after the elbow to line back up with your existing duct, which would probably eradicate all of the savings or even make things worse. And again, that would only be true if your duct velocity was relatively high compared to what I typically would associate with a residential system.

      You could get a sense of that by dividing the flow through the duct in cubic feet per minute by the cross-sectional area in square feet, which would give you the velocity in feet per minute. You could get a sense of the flow rate by multiplying the face area of the gills/registers the system serves in square feet by 300 – 500 feet per minute. That would give you a sense of what the flow rate was for each grill and if you added them up, the flow rate for the duct in question.

      I hope this is helpful to you.

      Best regards,

  2. Bruce Savage says:

    I calculated the flow rate for the duct and it is just under 750 fpm, just as you predicted. Using your register method (*400) I came up with around 300 fpm.

    One of the hvac guys around here said he has installed a vane in a similar tight spot commercial. He cut a curved slot in the accessible side of the duct with then pushed a heaviy gauge curved metal piece into the slot. The curve piece had an attached base plate that allowed him to tie the duct back together. I ask him if it helped and his response was that he didn’t know and didn’t care, it was what his customer directed him to do.

    I worked in a co-locate facility for some time and I was always amazed how the “internal weather” patterns changed when the CRACs KICKED IN. High air volume really changes things. Based on what you said above I think this addition would “in vane” so I will stay away from the sheet metal work. I really prefer keeping the skin on my hands and blood in my body.

    Thanks you for your response,

  3. Daygee says:

    thanks for the job here, but how can i calculate the number of turning vanes for this type of elbow.

    • Hi Daygee,

      Sorry for the delayed response; just catching up on things.

      I think if you really wanted to, you could probably come up with some sort of methodology for coming up with the exactly right number of turning vanes for a given elbow at a given flow rate with a given turning vane design. But the way we actually do it in our industry is to select from standard designs that have been developed and evaluated by vendors and industry organizations.

      I personally tend to use the ASHRAE duct fitting loss coefficients. When I first started doing it, they were in the form of tables that you could reference in the back of the fundamentals handbook. I think the last handbook where they were included was the 2009 Fundamentals edition.

      Here is an example of what that looked like for one of the radiused elbows like discussed in the first post about elbows with turning vanes.

      For complex geometries, you had to go through a number of steps and tables to come up with a loss coefficient.

      But for simple fittings, and mitered elbows fell into that category, there was a single number you used no matter what the dimensions of the elbow were. Here is what the handbook data looks like for one of the mitered elbow designs.

      In the 2009 handbook, there were 4 additional mitered elbow designs along with the one above, including one with no turning vanes.

      As you can see, the differences are related to if the vanes are single or double wall and differences in the distance between the vanes and the specific dimensions of the vanes themselves.

      When the information was first published – my recollection of first seeing it was in about 1980 or so – I was surprised to realize that those details meant that the pressure drop for a mitered elbow could vary by a factor of about 4, depending on the details of the turning vane design; so mitered elbows with vanes are definitely not all created equal!

      I suspect others might also be surprised to discover that because most specs I look at just ask for turning vanes but do not get into the specifics of the details of the vanes. And that can make a difference.

      One project a while back, the designer did specify the vanes in detail, but the spec did not get enforced and the result was a system that had a significantly higher static requirement than was anticipated by the design. On an elbow by elbow basis, the difference was not large. But with multiple elbows at high velocities, it can add up.

      At one point – about 2002 I think – ASHRAE developed a software based electronic database that automated the manual assessment process. In other words, you find the fitting of interest in the database, enter the basic dimensions and the flow rate, and the database comes up with the loss coefficient vs. you having to do the math to enter the tables. That database includes one additional double wall turning vane design that has a loss coefficient of 0.38

      Since the fitting loss is a function of the loss coefficient and the square of the velocity pressure, the significance of an improved loss coefficient will go up non-linearly with the velocity in the duct.

      In other words, a mitered elbow with no turning vanes in it would not be that big a deal at low the low velocities you might find in small ducts in the distribution system downstream of a terminal unit, where the duct velocities may be less than 1,000 fpm, maybe even only 500 – 600 fpm for a small duct.

      In contrast, a mitered elbow with no turning vanes in it in a large duct where the velocity was up in the 2,000 feet per minute range or higher could be a total disaster in terms of pressure drop, especially if your design calculation had anticipated that there would be turning vanes. And at those velocities, the differences between the different turning vane designs can add up.

      So I guess my bottom line in response to your comment is to suggest that you select a turning vane design from one of the options that have been tested by ASHRAE based on the nature of your duct system (high vs. low velocity) and then write your spec to ensure that you get that design for the installation. Asking for submittals on the turning vanes and then spot checking the installation during construction will probably be an important element of success because I am somewhat convinced that not everyone realizes that there really are differences in the losses associated with different turning vane designs.

      I hope that helps, and thanks for visiting the blog.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s