|Project by rodneyh||posted 02-24-2012 08:05 AM||3641 views||25 times favorited||12 comments|
Finished my mobile downdraft table. Same construction as my recently posted mobile sanding station. It’s 48” x 25”, and 36” tall. The top is split into 3 sections, each being about 13.5” x 19.5”. Each of the sections has a lip to hold a piece of 1/4” pegboard (for now). For smaller projects, 1 or 2 of the pieces of pegboard can be replaced with plywood (as in the 1st pic). It’s got ample storage, with 4 large full depth drawers and a top wider drawer. The top one is sized to accomodate the sloped downdraft section. I’ve only briefly tested it by hooking it up on a short run to my Jet DC100C, but it definitely eliminates most of the visible dust. I’ve got lots more optimization to do (see below), but it’s a great addition to my shop as is.
The following is a somewhat technical discussion of downdraft table (DT) design and how I plan to optimize my own. While I haven’t done work specifically on DT, I am a Mechanical Engineer with a PhD in Fluid Mechanics. Not trying to hype, just letting you know I’m not pulling this (at least not all) from my arse.
First, the requirements – we want high velocity (U) above the surface, large flow rate (Q) thru the system, and have the flow uniform across the surface. The entire design optimization is driven by the trade-off between uniform flow and velocity / flow rate.
1st let’s look at the flow path in real general terms. I’ll discuss it as if I were blowing air in thru the blast gate, as I think it’s easier to understand. It’s (nearly) the exact same thing in reverse when we start sucking air out the gate. Flow comes in thru the blast gate, flows across the length (X direction) of the table via a 10” square(ish) duct, and thru a slot into the upper V-shaped chamber. From there it moves primarily to the front and back of the table (Y direction), and then out thru the holes in the pegboard (Z direction).
Now let’s take a closer look at distributing the flow in the X direction. This is done via the duct and the slot. The duct is straight foward. You want this to have minimal pressure drop to improve all of the requirements. This is accomplished essentially by making it as large as possible. The 10” square duct for mine could have been doubled if the top drawer was eliminated. The effective length of the duct could have been cut in half by placing the blast gate in the center top back of the cabinet. For some designs, these improvements may be necessary. For mine, however, the huge 10” square duct has such little pressure drop that these steps weren’t necessary. The slot height gets a bit trickier, because there is definitely a trade-off between flow rate (Q) and uniformity. Think of it like this: If I make it too small (a few mm), it will have high pressure drop and reduce Q – that’s bad. If I make it too big (a few inches), all of the flow coming thru the blast gate will go out that large slot near the left end of the table yielding poor uniformity – that’s bad as well. We want to be in the ballpark of the slot having 10X the pressure drop that the duct does. This will ensure that we get flow thru the right hand side (RHS) of the slot that is about 90% of the flow out the LHS. If you increase the slot height to lower the pressure drop to 5X the duct, you’ll only get 80% out the RHS. Recall that the duct pressure drop should be extremely low, so having 10X that is not necessarily all that bad. For complicated geometries such as these, you can’t calculate pressure drops and velocities very accurately. That said, I winged it to get in the ballpark. I’ll go back later with an anemometer (air velocity measurement tool) and adjust the slot to give (at most) a 10% flow difference between the LHS and RHS of the slot.
From the slot to the holes in the pegboard is a very similar trade-off as the duct/slot trade-off discussed above. If the holes are too small, you’ll get poor Q (though you may still have high U near the table surface in the vicinity of a hole), and if the holes are too large (maybe a couple inches), you’ll have good flow only at table surface directly above the slot (e.g. poor uniformity). For now, I’ve just installed standard 3/16” pegboard with 1/4” holes on 1” centers. I’ll optimize the hole size similarly for uniformity and Q via anemometer measurements at various heights above the table surface.
The real “customer” requirement – The customer of this thing (me), does’t really care about velocities, flow rates, uniformity, or even the amount of dust at the table surface. What I really care about is the quantity of dust in the air above and in front of the table at a height from about 4’ to 6’, as this is where my head will be. I have access to some pretty sophisticated equipment to measure this. Once I get part way thru my flow optimization, I’ll figure out a way to quantify my improvements in terms of dust quantity and (hopefully) particle size.
Aside – In addition to optimizing the downdraft table itself, I believe there is a lot of opportunity for improvement based on where the unit is placed in the shop. Think of 2 cases: In the “baseline scenario” the table is in the middle of the shop. In this case, there is no appreciable help from the air currents in the shop. In the “improved scenario” the table is in a corner of the shop, and there is a curtain (floor to ceiling, maybe 12’ long) along the other side of the table, such that the table is at the end of a hallway. You’d suck air thru the table at the end of the hallway and pipe it to the DC somewhere in the main part of the shop. The air you sucked out of the hallway is replaced with air from the shop, effectively turning the hallway into a (very) low speed wind tunnel. This should keep any particles floating in the air moving in the general direction of the table. This is especially true for the smaller particles, as they will track the airflow much better than larger particles.
I realize that’s a pretty deep dive into this for most. Hopefully at least a few find it interesting. I’ll try to update this once I get some measurements in a few weeks.