In-Duct Air Velocity?

[DanCostello]

So, I've seen plenty of guidelines regarding air velocity at the HVAC registers (<300FPM at least. Ideally lower), but I've seen far fewer suggestions about limits for the velocity of the air within a duct. I found this post from Stuart suggesting that 675 FPM is okay. But I'm hoping to confirm that and maybe get a little bit more context for it, e.g. where the turbulence from the airflow starts to become problematic.

Getting a good register size isn't really a problem for me. But the peculiarities of my house and its myriad ancient systems are pushing me towards running the new ducts through some smaller-than-ideal passageways (so. much. plumbing.) and probably through at least one joist, so the smaller I can get some of this ductwork, the better.

And speaking of running through a joist: My vocal booth needs about 43 CFM of air and at least one of the ducts (probably the supply duct) is going to have to run through at least one of its 2x10 ceiling joists. At a duct speed of 675 FPM, that would require a round duct with a diameter of 3.42". Round that up to 4" and that's too big a hole to put through a 2x10. But I was thinking that I could run a 6" duct (lined, so the inner diameter is ~4") to the room, then just before it hits the joist, split that 6" duct into a couple 3" (or smaller) ducts, run those smaller ones through the joist, and recombine them in the silencer on the other side.

Does that sound not terrible?

-Dan.

[Soundman2020]

Welcome to the forum, Dan! Glad to see you here.

OK, air velocity in the duct: With two caveats, it really doesn't matter too much. The first caveat is: as long as we are talking about ducts that are OUTSIDE of the inner-leaf. In other words, ducts where the air flow goes through a silencer box that separates it from the register, inside the room. So, for example, you might have an AHU outside the studio that is pushing out air at maybe 600 FPM, which then goes to a duct that has a smaller cross section, so the velocity goes up to 800 fpm (for example), then it hits the outer-leaf silencer box where the internal cross section drops that down to, eg, 400 fpm... but then it goes into a duct that runs between the two leaves, and the velocity increases up to 800 fpm again.... then it reaches the silencer on the inner leaf, where the internal cross section drops it down to 300 fpm, and the final register is even larger, where it comes down to 250 fpm. That would be a valid scenario, since all the "noisy" higher speed air flow is outside of the inner-leaf, where it doesn't matter: because there's a silencer in between the high speed flow, and the register.

The important thing is that the air flow velocity at the register should not be higher than about 300 fpm, because air flowing faster than that across the vanes of the register, can produce audible noise. Also, you want it slowed down to that speed a bit before it reaches the register, and hopefully going along a straight run for many inches before hitting the register, to ensure that there's as little turbulence as possible in the flow, because turbulence = noise, and turbulent flow hitting a register = even more noise.

The second "caveat" is that there are reasonable limits for the velocity inside any give diameter or cross section of duct: Because: Friction loss! The higher the air velocity, the more energy it loses as it moves due to friction with the duct walls. In fact, friction loss is basically the same as aerodynamic drag, which increases according to the SQUARE of the velocity. So if you double the velocity, you get FOUR TIMES the drag, and if you quadruple the velocity you get SIXTEEN TIMES the drag. Also, the higher the velocity, the higher the turbulence. The general rule of thumb is to keep the air flow velocity below about 1000 fpm inside any duct, for studios. Commercial buildings have much higher flow velocity: maybe 2000 fpm, but that's too high for our purposes. 1000 fpm is already high, and for the size ducts we use in home studios, the losses are starting to get pretty high.

Here's a chart that might be useful in figuring all this out: different flow rates, duct sizes, and flow velocities, to keep things simple.

HVAC---air-duct-sizing-graph-flow-rate-vs-velocity-vs-duct-size-2.png (75.52 KiB) Viewed 45845 times

HVAC---air-duct-sizing-graph-flow-rate-vs-velocity-vs-duct-size-2.png (75.52 KiB) Viewed 45845 times

So, for example, your target here for the register, is the dark green colored line labeled "300 fpm". Never use any combination that would go to a LOWER line on the chart, but a higher line is fine. So the light green line at 250 fpm is fine, but not the light blue line at 350 fpm. So, using the light green line, you can see that if your room needs a flow RATE of 200 CFM, then the register size needs to be about 0.8 square feet or greater. And if you really wanted to keep things extra quiet, you could use the orange line (150 fpm), which shows you need a register area of nearly 1.5 square feet. And use the same chart for internal duct sizes: if you don't want things going faster than, for example, 600 fpm, then use the yellow line. Etc. You can use the usual equations to arrive at the same conclusion, but the graph makes it easier for people who don't like math!

In general, I try to keep internal duct velocities under about 800 fpm, and under 600 fpm if I possible. Basically, the lower you get the flow velocity, the better it will be overall for efficiency of the system, air noise, and static pressure.

- Stuart -

[DanCostello]

Ok, cool. Thanks. I can't run the numbers until I get home, but those guidelines actually seem a lot more generous than I was expecting; and at the risk of jynxing myself, this might one aspect of the project that turns out to be a little easier than I feared. :-p

-Dan.

[Soundman2020]

At a duct speed of 675 FPM, that would require a round duct with a diameter of 3.42". Round that up to 4" and that's too big a hole to put through a 2x10.

Or you could use a rectangular duct for the portion where you need to get through the joist. For penetration of structural joists, the limiting factor is usually how much wood you are leaving below the hole (and also above, in certain locations). Stresses in the joist are at their lowest value in the center of the joist and towards the top, and much higher down near the bottom edges. So you can often have a longer but narrower hole, rather than a round hole or a "fat" hole. You could possibly even have a quite large rectangular notch cut in the upper face of the joists, as long as it does not exceed the limitations allowed by your local building code, and as long as it is in the area of the span closest to one of the ends (within 1/3 of the span). The image below shows one such guidance (which may or may not be applicable to YOUR location! Check with your local authorities...):

So, assuming this is applicable in your area, you could have a notch up to 1-1/2 deep and 3-1/16 long in the top edge of your joist, through which you could put a rectangular duct. That would give you a cross-sectional area of about 4-1/2 in2. Two such notches would total 9in2 area. That might be enough. But don't take my word for it! Check with your local authorities, or get a structural engineer involved. Don't take chances when dealing with structural members in your studio.

then just before it hits the joist, split that 6" duct into a couple 3" (or smaller) ducts, run those smaller ones through the joist, and recombine them in the silencer on the other side.

Careful with that! Two 3" ducts is not the same area as a 6" duct! The cross sectional area of a 6" duct is about 28.3 in2, but the cross section of a 3" duct is not half of that! In fact, it's only one quarter: about 7 in2. So you would need to split a 6" round duct into FOUR x 3" round ducts if you wanted the same cross sectional area. Don't forget that there's a "squared" term in the equation for figuring area from diameter. Area= (PI/4) × D^^2. Or Area=PI x R^^2 in the case of using the radius instead of the diameter. Thus, cross-sectional area increases proportional to the square of the radius, or of the diameter.

But I was thinking that I could run a 6" duct (lined, so the inner diameter is ~4") to the room,

In normal HVAC nomenclature, that would be considered a 4" duct, not a 6". It's the internal diameter that matters, regardless of how much insulation it has on the outside. The cross-sectional area of a 4" duct is about 12.5 in2. So with three of those notches that I mentioned above (separated by at least 2", and only in the outer 1/3 of the joist), you could have about the same area as a 4" duct. If you wanted to reduce from a 4" duct to a 3" duct, the area would decrease from 12.5in2 to 7 in2, which is very roughly 50%. So the velocity would be very roughly double (OK, it would be 78% higher, to be exact, but call it double to be safe).

Yup, it can get complicated! Lots of stuff to consider...


- Stuart -

[DanCostello]

At a duct speed of 675 FPM, that would require a round duct with a diameter of 3.42". Round that up to 4" and that's too big a hole to put through a 2x10.

Or you could use a rectangular duct for the portion where you need to get through the joist. For penetration of structural joists, the limiting factor is usually how much wood you are leaving below the hole (and also above, in certain locations). Stresses in the joist are at their lowest value in the center of the joist and towards the top, and much higher down near the bottom edges. So you can often have a longer but narrower hole, rather than a round hole or a "fat" hole. You could possibly even have a quite large rectangular notch cut in the upper face of the joists, as long as it does not exceed the limitations allowed by your local building code, and as long as it is in the area of the span closest to one of the ends (within 1/3 of the span). The image below shows one such guidance (which may or may not be applicable to YOUR location! Check with your local authorities...):allowed-notches-and-hole-drilling-in-joists-2.jpg So, assuming this is applicable in your area, you could have a notch up to 1-1/2 deep and 3-1/16 long in the top edge of your joist, through which you could put a rectangular duct. That would give you a cross-sectional area of about 4-1/2 in2. Two such notches would total 9in2 area. That might be enough. But don't take my word for it! Check with your local authorities, or get a structural engineer involved. Don't take chances when dealing with structural members in your studio.

Yeah, I'd thought about doing a rectangular hole (and those numbers are, indeed, applicable to my area) - but the codes don't really cover a hole that isn't a notch in the top face, and trying to deal with my city codes people on something even slightly non-standard is more hassle than it's worth. But either way, it's much easier to make a 3" hole with a hole saw in the middle of a joist than it is to cut a 1.5"x3" notch out of the top of a joist that's currently holding up a concrete slab. :-p Fortunately, I have plenty of space to make multiple small holes.

then just before it hits the joist, split that 6" duct into a couple 3" (or smaller) ducts, run those smaller ones through the joist, and recombine them in the silencer on the other side.

Careful with that! Two 3" ducts is not the same area as a 6" duct! The cross sectional area of a 6" duct is about 28.3 in2, but the cross section of a 3" duct is not half of that! In fact, it's only one quarter: about 7 in2. So you would need to split a 6" round duct into FOUR x 3" round ducts if you wanted the same cross sectional area. Don't forget that there's a "squared" term in the equation for figuring area from diameter. Area= (PI/4) × D^^2. Or Area=PI x R^^2 in the case of using the radius instead of the diameter. Thus, cross-sectional area increases proportional to the square of the radius, or of the diameter.

Sure, sure. That's what I get for not being precise in my wording. :-p

-Dan.

[DanCostello]

Am I correct in understanding that duct liner is really only needed in the ducts on the inside of the inner leaf? Or is it needed in all the ducts everywhere?

(This is aside from the mufflers, which would all have duct liner)

-Dan.

[Purelythemusic]

From my super limited knowledge and soon to be moderately limited experience, I think I can say it is not necessary outside the inner leaf... where are your ‘inner leaf silencers’? If they’re outside the inner room then just the duct/sleeve from the inner room silencers would need liner... I’m sure a more certain answer will come

[Soundman2020]

Tom has the right idea.

But what type of duct are we talking about? If it is Flexduct, then it doesn't need extra insulation. If it is sheet-metal duct, then like Tom said, it would be a good idea to insulate it for the sections between the two leaves, and if you are also using an HRV or ERV (in addition to the actual AHU), then it would be best to insulate the duct runs between the ARV/HRV and the AHU, for maximum efficiency. The sections beyond that don't really need it.

Also, there are two completely different types of "duct liner", which should not be confused: There's the proper duct liner for the interior of ducts, which is designed to withstand the movement of air over its face ad has both thermal and acoustic properties, but there is also "duct wrap", which is meant to go on the outside of the duct, and is mostly for thermal insulation, not so much for acoustic isolation. Don't mix them up! In some parts of the world, it can be hard to find proper duct-liner (ask Tom about that! ), and some less-than-honest suppliers will try to sell you duct-wrap, insisting that it is duct-liner... Caveat emptor....

- Stuart -

[DanCostello]

My most immediate concern is the "cold" air returns from the vocal booth and control room. (I'll deal with the supply ducts later. I have a bit more wiggle room on placing and sizing those) My current plan is for the vocal booth & control room air returns to each have their own inner leaf silencers, built into the walls. Ductwork leading from these silencers will be made from sheet metal. The ducts from the two rooms will be joined together with a wye connector between the inner and outer leaves (i.e. inside an exterior wall), and the return air will flow through a single duct to a single silencer on the outer leaf.

And I'm specifically talking about duct liner that goes inside the duct, not thermal duct insulation. Not only do I not really care about the return duct losing thermal energy, but it's going to wind up surrounded by rockwool eventually anyways.

Since we're on the subject, the combined airflow from both rooms warrants a duct cross section of about 45 square inches. Assuming I don't need to add duct liner to this run, is there any issue with using 3.25"x14" rectangular "stack duct" for part of the run rather than just 8" round duct for the whole thing? The cross section numbers work out, but I wasn't sure if the narrow profile would cause the static pressure to increase such that I'd need an overall larger duct.

If I do need to add duct liner to these returns, then stack duct is off the table because it would be almost entirely filled with the liner.

-Dan.

[Soundman2020]

the combined airflow from both rooms warrants a duct cross section of about 45 square inches. Assuming I don't need to add duct liner to this run, is there any issue with using 3.25"x14" rectangular "stack duct" for part of the run rather than just 8" round duct for the whole thing? The cross section numbers work out, but I wasn't sure if the narrow profile would cause the static pressure to increase such that I'd need an overall larger duct.

Unfortunately, there's not really a direct comparison between the areas of round duct and rectangular duct. It would be nice if it really did work out like that, but Murphy will always get you!

Here's a table that shows the actual equivalence between the two types of ducts, based on friction losses.

https://www.engineeringtoolbox.com/equi ... d_443.html

As you can see, the equivalent of an 8" round duct would be 7x8, 6x10, or 5x12. That's 56 in2, 60 in2, and 60 in2, respectively, as compared to aprox. 50 in2 for the 8" duct. Rectangular ducts need to be larger than you think to get the same flow as round ducts. Oval ducts are different again... (see here: https://www.engineeringtoolbox.com/equi ... d_205.html )So yours would need to be maybe 3.25 x 17 or even 18 to get the same friction loss as the 8" duct.

Now, that said, you could still use 3.25 x 14 as long as it is a fairly short section, and as long as your AHU is able to handle the total static pressure of your system (including all of the ducts, silencers, elbows, dampers, registers, etc.

These might also be useful in helping you figure things out:
https://www.engineeringtoolbox.com/duct ... d_444.html
https://www.engineeringtoolbox.com/equa ... _1028.html

Isn't HVAC fun?! Nothing is as simple as it seems.... Sigh!


- Stuart -