# Measuring bean speed

E

#### Ed Mulligan

We sometimes move coffee beans by pneumatic conveying (vacuum). Is there a way to directly measure the speed of the beans as they travel through the system? I need to do this without actually touching the beans. Thanks, Ed Speaking for me, not for Starbucks. . .

K

#### Koen Verschueren

What you can try is using a ultrasonic flowmeter. I don't know if there is a big difference between the speed of the beans and the speed of the air and if this will give you measuring problems. Look at http://www.krohne.com for more information about the measuring principle of ultrasonic flowmeters. You can find these flowmeters from different suppliers. There are even some types that can be clamped on a pipeline.

J

#### Jon Kirwan

> We sometimes move coffee beans by pneumatic > conveying (vacuum). Is there a way to directly > measure the speed of the beans as they travel > through the system? I need to do this without > actually touching the beans. I assume you are open to an optical method. If so, you might consider using correlation. I'm imagining a tube of coffee beans with somewhat varying densities of coffee beans flying by any particular point. A defocused optical system could gather an average of light passing through an instantaneous "slice" of the coffee beans. This might be an array of detectors and emitters providing a linear slice or you might come up with a single average or a single narrow beam or the defocused system already mentioned. To keep it simple, it would be nice to have a simple scalar value at successive time points representing the "opacity" of the beans in flight at that point on the tube. A second system would be set up very near to the first. Keeping it close means that the shape of the "bean cloud" passing by at one point will not get a chance to significantly change shape before it passes by the second detector system. It's important that the rate of change of the shape of the "bean cloud" is slow compared to the rate of speed that the beans are traveling at (the phenomenon of interest in this case.) If you take a sequence of data points at the first sensor site and a sequence of data points at the second sensor site, the only substantial difference in the two time-amplitude signals should be a time delay. This time delay is the time it takes for the "bean cloud" to travel the distance between the sensors. Cross-correlating the two signal streams should produce that time delay. Knowing the distance, you should be able to calculate a speed for the beans. The shape of the cloud will be largely irrelevant for this process. Of course, I don't really know much about your needs, so I'm really guessing a lot. But problems like yours are sometimes solved using these correlators. Feel free to write. I'm at [email protected]. (I just visited this site for the first time today and I'm not sure whether I will come back again, or not.) Good luck, Jon

J

#### jmGiraud

> We sometimes move coffee beans by pneumatic conveying (vacuum). Is there a way to directly measure the speed of the beans as they travel through the system? I need to do this without actually touching the beans. < ____________________ Edward: Could contact Alesa Alusuisse: They are specialist in moving Aluminium oxide in air conveyors. Works perfect. They may have good idea how the speed of beans relate to the flow of air ? does it make sense ? Could you relate the weight of the transported beans and the time and the pipe size ...

G

#### Gilles Allard

Hi! You do not specify if your flow measurement is required only for a verification or if you need it for continuous operation. Let me suggest this approach. Direct your bean flow to a weighted vessel. Record the weight of the vessel and the vacuum pressure at the bean introduction point. The slope (derivative) of the weight is your flow. You can correlate this with the pressure and thereafter you can use the pressure to indirectly determine the flow. This will only give you an approximate measurement but since you do not specify the accuracy required, it may be enough. Please note that pressure vs flow curve will vary with each bean type (and size). You may have to do many calibrations if you normally carry many products. I hope this help. Gilles

R

#### Rick L. Hudson EMCO Inc.

Have you looked at an acoustic based doppler method. Some have been used to measure fluids in pipes. Rick Hudson

J

#### Jeffrey W. Eggenberger

It should be a function of the pressure differential in the system. The pressure differential should be easy to measure. Jeffrey W. Eggenberger Electrician: Industrial, Commercial, and Residential