Between a drawing and a machine that can leave the factory there are a number of specific operations. Most of the problems customers meet on site can, if they are traced back, be traced to a trade-off made in one of those operations. The sections below take a few of the critical operations in order: what we insist on there, why we insist on it, and where the cost lands if it is left out.
Sheet metal and machining
The sheet metal parts are not only about shape. Inside the oil tank, in the vacuum chamber and in the drip tray the sealing is involved, and the quality of the weld seam decides directly whether a vacuum can be established later. The plate thickness, the welding distortion and the flatness of the mating surfaces all show up in the leakage rate of the vacuum system. Cutting corners in this operation is often hard to see: there is no visible difference in the appearance, but the pressure comes back up noticeably faster during the hold-up test.
For the machined parts the focus is on the mating faces and the pipe connections. The flatness of the flange faces, the quality of the threads and the chamfer and the end-face treatment of the oil pipe connections all turn into specific torque requirements at the assembly stage. A connection that is short of accuracy may feel tight during assembly and will start to weep after it has been taken apart and reassembled two or three times, and a leak point of that kind is very hard to locate during operation, so the whole section of pipework has to be replaced.
Assembly of the oil circuit fittings
The problem that occurs most often in the oil circuit assembly is that something is fitted the wrong way round. A valve fitted in reverse, a pipework connected incorrectly, a seal left out: none of these is easy to find before the machine is energised and the oil is flowing, and by the time they are found the work has already been wasted. Checking the direction of the flow and the orientation of the valves section by section against the drawing is the basic requirement of this operation; every arrow on the drawing corresponds to an actual direction of movement.
Another point that is easily overlooked is cleanliness. The fittings should be clean before they go into the machine, and if debris and fragments of the sealing material fall during assembly and are left in the pipework, the life of the filter elements is shortened directly. Skipping the cleaning operation saves a few minutes of labour and costs the life of the first batch of elements on the customer's site; that account is usually only settled after the delivery.
Electrical wiring and cabling
The electrical wiring is done to the drawing, but what really has to be watched is whether the operation of the protections and the interlocks has been verified item by item. The overpressure alarm with sound and light that stops the machine, the phase loss and phase sequence protection, the interlock with the automatic defoaming: once the assembly is complete, each of these functions has to be triggered and verified one by one. Looking only at the panel is not enough, because a correct display does not mean that the action is correct.
The cabling also has to take the later maintenance into account: whether the wire identification is clear, whether the control cables and the power cables run separately, and whether the tightening torque of the terminals is consistent. None of this affects how the machine performs as it leaves the factory, but it decides how quickly faults can be found on site. A machine that is neatly wired takes much less time to repair, and it is also harder for a customer dealing with a minor fault to connect a wire wrongly.
Leak testing of the vacuum system
This is the operation that shows the assembly quality most clearly. The leak test is usually done by staged hold-up: first hold the whole machine up to see the overall leak rate, and then isolate it section by section to locate the fault in a specific pipe section, joint or chamber. The ultimate vacuum of ≤5 Pa is the best level the machine can reach, while in actual operation with oil the working vacuum stays in the -0.08 ~ -0.0999 MPa range. Both depend on the leak test.
If the leak test consists of a single overall hold-up without the staged locating, the leak is often not found, because the curve produced by several small leaks together is very much like that of a single large leak. The staged leak test takes longer, but it gives a definite position for the rework, it avoids taking the whole machine apart repeatedly, and it lowers the cost of the fault tracing that goes with the delivery as well as the waiting time on site.
The oil test run after assembly
The last gate is the running-in with oil. At this stage the observation is not only whether the machine runs at all: whether the noise of the pump is normal, whether the inlet and outlet pressures are steady, whether the rate at which the differential pressure across the element rises is reasonable, whether oil weeps at the seals, whether the temperature control holds within the set range, and how the vacuum is established and maintained. These observations are written into the test run record and are delivered with the machine.
A machine that has only been run with no load can expose problems the first time oil is admitted on site: the mechanical fit of the pump, the seating of the seals, the residues in the oil circuit, and none of these can be seen until it has actually run. The oil used for the test run is recovered afterwards, so this step is also a rehearsal of the cost of processing oil on site; the results can be compared directly and the data from the two sides line up easily.
| Operation | How cutting corners shows up | How it shows up on the customer's site |
|---|---|---|
| Sheet metal welding | The weld seam has porosity or has not been fully penetrated | The vacuum level does not come up and the pressure comes back up too quickly during the hold-up test |
| Fitting assembly | A cleaning operation is skipped and the seals are left out | Short filter element life and fluctuating pressure |
| Electrical wiring | The protective devices are connected but never tested | The machine trips on site and the protection does not operate |
| Vacuum leak test | Only the overall hold-up is carried out | The leak cannot be found and the result of the treatment is not stable |
| Oil test run | Only a no-load run is carried out | Noise, oil weeping and an abnormal differential pressure appear as soon as oil is admitted for the first time |
Every step in the workshop is eventually magnified on the customer's site.