# How to measure absolute zero vacuum in pumps

In a vacuum pump, does the inlet pressure measure closer to 30 inHG?

The following is a pure water pump, a 100w 8 meter regenerative turbine pump. Not a vacuum pump. Maybe the reason it couldnt get lower than 20 inHG was because its not a vacuum pump? Or is the analysis the same in both? How low can a real vacuum pump measure in the suction side?

I wanted to understand the physics side of vacuum in pumps, so I bought a compound pressure gauge that can measure absolute zero to atmospheric pressure and positive pressure.

I throttled the ball gate valve to decrease atmospheric pressure in the inlet of a 100W Regenerative turbine pump.

But the most I could measure is -20 inHG. I can't get to -30 inHH or absolute zero. Will it work if I move the pressure gauge closer to the impeller? But as I understand, vacuum is uniform in enclosed space, or does it vary in different portions inside? I can't find any adaptor that can make the gauge closer to the impeller. I have to let fabricator design one and it will be expensive and will take long.

If I used a centrifugal pump, will the vacuum be absolute zero close to the impeller? I assume the eye of the impeller is absolute zero vacuum? how large does the vacuum extend and how do you compute it (if you know how)?

• Serious vacuum requires two or more pumps in series and a cold trap. – blacksmith37 Oct 24 '20 at 14:02
• what is a cold trap? you mean it's not even absolute zero vacuum in the eye of the impeller? – Jtl Oct 24 '20 at 14:05
• As per my comment on my own answer below, if the pump is meant for water, the designers were probably trying to prevent any absolute pressures below about $0.15\,\mathsf{bar}$ being produced anywhere, so in trying to get a hard vacuum, you're working against the pump designers. The lowest absolute pressure I ever managed to produce with "real vacuum pumps" was $8\times 10^{-14}\,\mathsf{bar}$; but I believe there are people who are much more careful about keeping their cold traps topped up with liquid nitrogen than I ever was, and they can probably get lower. – Daniel Hatton Oct 27 '20 at 15:08
• I meant an absolute pressure of $0.15\,\mathsf{bar}$ or $4.4\,\mathsf{inHg}$, which is the same as a gauge pressure of $-0.85\,\mathsf{bar}$ or $-25.6\,\mathsf{inHg}$. In the photo, your meter is reading a gauge pressure of $-16\,\mathsf{inHg}$, which is above $-25.6\,\mathsf{inHg}$. – Daniel Hatton Oct 27 '20 at 20:41
• Tricky. The first thing that springs to mind is an oil-sealed rotary vane pump, but they cost of the order of thousands of US dollars and, worse, vent carcinogenic oil vapour on the atmospheric pressure side, so you'd need to install a proper fume-handling system. Second thing to spring to mind is a zeolite sorption pump. You could probably construct one of these yourself from cheap materials, but there would still be significant safety challenges in mitigating the risks of burns, asphyxiation, and embrittlement of insulation on mains cables due to the presence of liquid nitrogen... – Daniel Hatton Oct 28 '20 at 12:49

There's a characteristic number for pumping processes called the "Thoma cavitation parameter". Although it has "cavitation" in the name, it's still meaningful irrespective of whether or not there's not a risk of cavitation (in your setup, I think there's a very high risk of cavitation). The Thoma cavitation parameter is defined (assuming that the inlet and outlet of the pump are at the same height and have the same fluid density and pipe diameter, which all seem pretty safe in your setup) as

$$\frac{p_{\textrm{i}}-p_{\textrm{l}}}{p_{\textrm{o}}-p_{\textrm{i}}}$$,

where $$p_{\textrm{i}}$$ is the pressure at the pump inlet, $$p_{\textrm{o}}$$ is the pressure at the pump outlet, and $$p_{\textrm{l}}$$ is the lowest pressure encountered anywhere in the pump.

AIUI, you want to compute $$p_{\textrm{l}}$$, and you already have an instrument measuring $$p_{\textrm{i}}$$. You can set $$p_{\textrm{o}}$$ to atmospheric pressure with decent accuracy by opening to atmosphere immediately downstream of the thing in the top left of your picture that I'm guessing is an ultrasonic flow meter. Hence, if you can find a value for the Thoma cavitation parameter, you've got all the ingredients you need to compute $$p_{\textrm{l}}$$.

The Thoma cavitation parameter depends on the shape of the impeller, so your best hope for finding a value is that the pump manufacturer has measured it. If you're lucky, the Thoma cavitation parameter will be a constant for that impeller shape, and the manufacturer will just be able to tell you a single value. If you're a bit less lucky, the Thoma cavitation parameter for that shape of impeller might depend on the values of two other characteristic numbers of the pumping process, called the "specific speed" and the "specific diameter", in which case what the manufacturer will (hopefully) send is a graph or table of how it depends on those numbers. If that really is an ultrasonic flow meter in your photo, then you'll have enough data to work out the values of specific speed and specific diameter, and look up the Thoma cavitation parameter on the graph.

• Its not ultrasonic flow meter but a flow sensor..i adjusted different discharge pipe even without one but none can get it below 20 inHG. – Jtl Oct 25 '20 at 1:15
• I wasn't proposing a way of making the pressure any lower, just a way of estimating what the lowest pressure anywhere in the pump is. – Daniel Hatton Oct 25 '20 at 8:47
• Does what you described apply also to regenerative turbine pump. Note "The main difference between a turbine pump and a typical centrifugal pump is its impeller design. Compared to most centrifugal pumps, turbine pumps have smaller diameter impellers with rows of numerous small vanes. These vanes recirculate the fluid as it travels from the suction end to the outlet. Specifically, fluid enters at the edge of an impeller blade (not through the eye) and is accelerated not only tangentially in the direction of rotation, but also radially outward into the casing channel by centrifugal force." – Jtl Oct 25 '20 at 9:24
• @Jtl Yes, it still applies. I'd expect the value of the Thoma cavitation parameter to be different for different types of pump, but that doesn't change the fact that one can define a Thoma cavitation parameter and use it to estimate the lowest pressure anywhere in the pump. Incidentally, if this pump is marketed mostly for pumping water, I'd expect the manufacturer to have made considerable efforts to make sure the absolute pressure doesn't go lower than about $0.15\,\mathsf{bar}$ anywhere in the pump, because going lower than that would mean cavitation that could damage the impeller. – Daniel Hatton Oct 25 '20 at 10:11

Practically speaking, I don't think any vacuum setup will get to an absolute vacuum because any connection point (flanged, threaded, etc.) will allow small amounts of air to leak in. Even when I worked with high vacuum chemical processes, the pressures we achieved were always 10's of millitorr (approx. 10$$^{-8}$$ inHg). A inHg gauge isn't going to give you the resolution you need to see that low of a pressure - and that's assuming the gauge calibration is good.

A pressure gradient has to exist in the connected pipe for there to be flow, so the closer to the pump, the lower the pressure - however, I don't expect the difference to be noticeable on that gauge because of its resolution.

For practical purposes though, many applications can be satisfied with 19 in.Hg vacuum, provided the rest of the system is okay.

• I wonder if your answer assumed it was a vacuum pump. What is the difference between a vacuum pump and regular water pump anyway? could they be the same with the former having tigher tolerances? – Jtl Oct 26 '20 at 1:31