4.3 Corrosion in pumps at high flow velocities
Corrosion in pumps at high flow velocities
There is a choice of methods of testing corrosion in pumps at high flow velocities:
- pipe flow
- rotating cylinder, freely rotating or enclosed
- rotating disc, freely rotating or enclosed
- spraying a sample piece with a thin liquid spray
If the purpose of the test is to determine the probable corrosion rate then the test rig should as far as possible resemble the condition intended to be tested. This is so that the test shall take place at the same Reynold’s number, i.e. similar flow. Otherwise the result will only be a comparison between different materials.
When the corrosion test is intended for pumps, it will be advantageous to use an enclosed rotary disc at a speed such that the Reynold’s number Red = (u2 * d2) / ν is about the same as that for the pump impeller. The units in the Reynold’s number formula are:
u2 = peripheral velocity of impeller (m/s)
d2 = outside diameter of the impeller (m)
ν = viscosity of the liquid (m²/s)
Instead of rotating a complete disc of test material, small electrically connected samples can be set into the fixed wall opposite the disc, see figure 4.3a, which are then exposed to the same velocity as the rotating disc.
Figure 4.3a shows a corromatic test set with
A — test plate B — rotor C — test chamber
D — test with fixing
E — speed motor.
This positioning of the sample pieces has the advantage that electrical measurements be carried out simply, see figure 4.3b. Besides which, surfaces subjected to various liquid velocities can at least be of the same size. In a solid rotating disc, the larger part of the surface area is subjected to the highest liquid velocities, whereas in the pump the opposite is the case, because of the large surface area of the pump casing.
Figure 4.3b Electrical connection of test pieces subjected to various liquid velocities. A — amplifier and recorder
The test arrangement shown in figure 4.3a and figure 4.3b will resemble the corrosion process in a pump, since all the relevant liquid velocities are present at once, and because redistribution of metal losses due to galvanic currents will occur between parts of the same metal subjected to various liquid velocities.
The galvanic currents are among the items measured in corrosion testing, apart from the usual weighing, potential measurements and so on. If metal corrodes, there are actual electrolytic currents flowing between parts of the same metal surface subjected to differing liquid velocities. Thus, by recording the currents, it is possible instantaneously to check the occurrence of corrosion, and at the same time qualitatively to follow changes in the rate of corrosion, build-up of protective film of corrosion products and so on. Unintentional changes in the rate of corrosion can also easily be recorded.
Materials generating thick protective films require relatively long test times, and worthwhile laboratory tests can be difficult to carry out, since it is not easy to maintain control over the composition of the liquid over long periods of time. In these cases it is better to move the laboratory to the liquid rather than the other way round. When there are decidedly passive films generated normally on the test samples by contact with the air, it is best to remove this film, i.e. activate, in order to establish whether the passivity is restored in the actual liquid. After activation, passivity will be restored after only a short time, within hours, if it is going to be restored at all. If the general corrosion is uniform with time after a certain stabilizing time, this is confirmed by the fact that the electrolytic currents also are uniform. The test time is then regulated so that stable conditions apply during as great a part of the time as possible.
Figure 4.3c shows some examples of test results using the equipment in figure 4.3a.
Figure 4.3c Onset of corrosion in pumps as a function of flow velocity.