Bulk handling machines convey a volumetric rate Q (m3/hr) and a gravimetric rate W (te/hr). These two rates are related by the bulk density of the material being conveyed D (te/m3).

The volumetric rate Q (m3/hr) is given by the area conveyed A (m2) times the effective conveying velocity V (m/sec), then chuck in 3600 for consistency of time units. So you vary the volumetric rate by varying the area conveyed, or the conveying velocity, or both.

The gravimetric rate W (te/hr) is given by the volumetric rate Q times the bulk density D. Varying the volumetric rate varies the gravimetric rate. Of course, if your bulk density varies then so will your gravimetric rate.

In the case of screw feeders they are frequently quoted as having a specific volumetric rate of XXX m3/revolution. depending on the size (diameter).

So there you have it, the simplified introduction. But life isn't always as easy as this.

With regard to your illustration I think I'd use a slightly different arrangement using one variable speed screw feeder under the bunker outlet to control the throughput then feed to a second screw conveyor that actually does the gravimetric measuring.

No doubt others may have different ideas, but that's the joy of bulk materials handling. ■

Originally Posted by Skynet

We have a worm feeder as shown in the attached diagram. This consists of a tube suspended on loadcells. The product (fine granular steel particles, approx. diameter of each particle is 1mm) enters the tube at the top left side, and then moves through the feeder. The product then leaves the tube at the bottom right. The dimensions of the tube are approximately 5000mm x 400mm.

The worm feeder is driven by a variable speed motor, which is generally used at approx. 1400 rpm. This speed is reduced by a gearbox, with a 21:1 ratio, so that the worm is turning at approx. 66 rpm. However, if too little or too much product is being transferred, this speed can be adjusted. The function of the feeder is to move product from one side of the plant to a conveyor belt on the other side of the plant, so that the product can then enter a mill. The only variables known are speed and mass. When there is no product within the worm, and the motor is switched on, the loadcells are tared to read 0kg. As the product enters the worm, the mass in the worm increases to about 100kg. I’ve noticed that while the system is running, the mass changes all the time (e.g. from 90-100kg). This is mainly due to the amount of product entering the worm (from a silo) not remaining constant.

An electronic controller board was developed to record the mass and give a warning alarm if the feed rate is outside pre-determined specifications. The speed of the motor is entered into the controller by a keypad. The mass is recorded by the controller at regular intervals (e.g. every second).

There is no shaft encoder installed on the axis of the worm feeder, because the speed of the worm remains almost constant all the time. This speed will only be changed by the user, who will then enter the new speed on the keypad.

We need to know the following:

1.Does the length of the feeder affect the result ? What result? But not really the issue.

2.How can I convert the angular velocity (66rpm) of the worm feeder into a linear velocity (m/s) for the movement of product along the feeder ? This is indeterminate unless you know the bed depth in the tube and of course the flight pitch.

3.What is the definition and calculation for the displacement of a screw (I’ve seen this on the forum, rpm x pitch x area). What is the unit for this value ? Roughly the same as 2

4.How do I determine the Feed rate (kg/s or kg/hour) ? Measure the time to discharge a known quantity.

5.How do I determine the Total amount of product displaced over a specific time period e.g. a 12 hour shift (kg) ? See 4

6.I also need to know how to calibrate the system so that I can know that the mass recorded is accurate over time (e.g. by using weights placed on top of the tube while the tube is empty, and then running the system for a specified time). You will first have to provide variance.

I would really appreciate some help with these questions since I have been struggling for a very long time to get answers. Thank you very much for your help.

Yours is a straightforward I & C situation. You have 11% variance due to incoming surges. Regulate the inlet valve. Average the loadcell signals and feed the result back to a VSD on the motor, and the valve actuator. I suspect the inlet valve regulation was overlooked at the electronic controller development stage. ■

You might consider that a theoretical linear velocity is the pitch of the flighting times the speed in rev/min.

Looking at the pictures the machine in question would appear as a screw conveyor rather than a screw feeder. Normally a screw feeder, or a feeder section at the start of conveyor, would have an internal shroud converting the casing into a tube rather than a trough and the screw flighting of shorter pitch. Are these features present in your machine?

I notice also that the hopper bottom appears to suffer significant "hopper rash".

Here is a link to an illustration of a screw weighing unit.

http://="http://www.transmin.com.au/...cus.php?pid=8" ■

Screw conveyor v. screw feeder, the important point is that you need particular design features if you need to feed a controlled rate from a flood feed inlet as I outlined previously. You also need reliable flow from the hopper, something I would question as the hopper bottom has severe "hammer rash". You need to get inside the machine and see what you've got!

There also seem to be air injection points just above the hopper outlet. You need to be careful with air, the last thing you need is a fluidised bulk material flushing uncontrollably through the conveyor.

As matter of terminology, we always talk about screw flighting to describe the helical platework scroll normally attached to the centre tube on a screw conveyor.

Your calculation is correct in theory although there may be some inefficiencies to be allowed for.

You appear to have an intermediate outlet in the conveyor. I would be collecting the discharge for a period of time and weighing it to give myself a a reference point against which to check my calculations. ■

Welcome to the wonderful world of dynamic weighing. You appear to be attempting to build a crude mass flowmeter. Based on my observations here are some ideas to try:

1. Your raw weighing measurement is being overly influenced by the headload of the silo and flowability of the material. As evidenced by the severe case of bin rash, the inlet flow into the conveyor will be erratic and will thus affect the loading of the inlet loadcells. By disconnecting the inlet end loadcells you will create a fulcrum at the inlet, and effectively will only weigh half of the material in the conveyor.

2. To circumvent having to prepare multiple calculations for flowrate vs feedrate, you can alternately take a weighed catch sample. Assuming your loadcell influence problem is fixed, simply run the conveyor for X minutes and note the speed & loadcell readings. Plot the readings on a curve, because typically conveyors are nonlinear relative to speed. Can you offload to a large container that can be check weighed?

3. There are controllers available that are dedicated to instrument your application. We have one that is PLC based, otherwise a good proprietary unit is the Milltronics BW500. Using a pre-programmed controller will eliminate you from reinventing their weighing algorithms. Info: http://www.lesman.com/unleashd/catal...accubw500.html

Regards, Delmar Schmidt

Melfi Technologies Houston

www.melfitechnologies.com ■

Originally Posted by Delmar

Plot the readings on a curve, because typically conveyors are nonlinear relative to speed.

Is that right?

Any or all types of conveyors?

Have you measurements to back it up? ■

Screw feeder suspended on weighscales may be good arrangement. But simpler arrangement would be to instal a Sankyo type impact plate scale in the discharge chute of the screw feeder. Impact scales of this type are also manufactured by Schenck. Signals from scale can be used for controlling speed of the screw. ■

The battered hopper and reported symtoms clearly indicate a flow problem from the hopper. There is a lack of information about the product, size of hopper outlet and inlet features to the conveyor and screw. Is the slide valve fully open?. Is there a choke section in the casing at the conveyor inlet. The pitch is given, but what is the diameter of the screw? Is there any substance on the surface of the steel particles, are they smoothly round and what is the bulk density from a loose poured condition?

Replacing the bottom of the feed hopper with a mass flow section, fitting a feed choke and making the inlet region of the screw with a reduced pitch would go far to improving the stability of feed to the sy tem. ■