Leakage diagrams and tables, which you can find on the data sheets of gasketdata.org, have become the worldwide standard when it comes to the presentation of leakage characteristics. Nevertheless, the representation often causes question marks among the not yet so experienced users. This page is an attempt to explain the presentation. If you have any comments, please feel free to contact us, e.g. via the contact form.
In order to understand why the presentation of the leakage data was chosen in this way, it is helpful to first look at the test procedure for determining the leakage data. Each leakage diagram represents the data from only one test rig test with a single gasket.1 The course of the test rig test is shown in the diagram below. 2
Diagram 1: Course of test
The diagram shows the test sequence in the form of the surface pressure curve and the applied internal pressure over the test duration in seconds.
The test according to the EN 13555 standard defines the use of helium as the test gas at a pressure of 40 bar and measurement at room temperature. The pressure of 40 bar is deviated from for certain gaskets if the intended use of the gasket material is for lower or even higher pressures. For gaskets that have a wider range of applications, more than one pressure is used for testing (e.g. 10, 40 and 160bar), each as a individual characteristic. The standard also specifies surface pressures up to which the gasket is loaded in order to carry out a leakage measurement. The load is usually followed by unloading (in several stages) in order to measure the leakage there as well. The first surface pressure is mentioned as 5 MPa in the standard, the highest is 160 MPa or, if the QSmax of the gasket is lower than 160 MPa, a surface pressure which not not overloads it. The leakage measurement is done by either mass spectrometry, mass flow measurement or pressure drop method at the particular surface pressure level until the leakage is constant (<= 2% over 20 minutes).
In practice, the test is carried out by mounting the gasket in a test rig that is modelled like a real flange. However, the surface pressure is not applied with bolts but with a hydraulic cylinder and can be maintained and adjusted. The inner space, which is the pipe in real use, is first evacuated and then filled with helium up to the specified pressure. Between the outer edge of the flanges and the outer diameter of the gasket there is an additional seal, e.g. an O-ring, so that the helium escaping from the interior space via and through the seal can be collected and measured.
In the diagram it can be seen that the 5 MPa surface pressure level defined by the standard is not used as the initial load. This is quite common, as certain seals do not have sufficient sealing performance at such low surface pressures. In this case, therefore, the gasket was directly loaded to 10 MPa, the leakage was measured, the load was increased to 20 MPa, the leakage was measured again, the load was decreased to 10 MPa and the leakage was measured again. This is followed by loading to 40 MPa and unloading to 20 and 10 MPa, in each case with leakage measurement. Leakage data are not shown in this diagram for reasons of clarity. The measurement procedure was continued in this example up to 160 MPa. The QSmax of this seal should therefore be at least 160 MPa. The pressure curve shows that the 40 bar internal pressure was applied at the beginning and maintained throughout the test.
The data from the test sequence is processed programmatically and transferred to the following view. The diagram shows the course of the leakage over the approached surface pressure levels.
Diagram 2: Leakage diagram
The courses and points shown are colour-coded to enable and simplify the assignment to the values shown in the next table.
The so-called loading branch is shown in purple, the unloading branches are shown in green. The points in purple and green represent the applied surface pressures and the leakage measured at this point. There is no measurement data between these points; they are simply interpolated for the representation in the diagram. Shown in pink are the intersections with a tightness class (LN), e.g. L0.01, which (as a rule) result from the interpolation.
If only the loading branch (purple) is considered, one recognises a course over all surface pressures, which in the first diagram form the upper plateaus before an unloading takes place. The lowest surface pressure is the surface pressure at which the test starts with the first leakage measurement (here 10MPa), the highest is 160 MPa in this example.
The green unloading branches always start from the loading load and are usually continued to the lowest surface pressure. Unloading in this example starts at 20, 40, 60, 80, 100 and 160 MPa.
The diagram can be read and used in different ways: Assuming the mounting surface pressure for the gasket material shown is to be 60 MPa and the PQR value at room temperature is 0.5, there would still be 30 MPa surface pressure left after installation of the gasket it has adjusted to the flange surface. In this case, tightness class L0.001 would be met. If the requirement is that tightness class L0.0001 is to be complied with, a point must first be found on the loading branch which is already below the required tightness class and from which an unloading starts. In this example, this would be 80, 100 or 160 MPa. Assuming 0.5 for the PQR of this gasket, 40 MPa would remain after mounting with 80 MPa and a tightness class L0.0001 would be certainly met. Only if the surface pressure during assembly at 80 MPa would be below 20 MPa, the tightness class L0.0001 would be exceeded.
The following table shows the data of the diagram in values. On the left side, the tightness class (L) and the minimum surface pressure during assembly (Qmin(L)) are entered. Horizontally, the minimum surface pressure during operation (QSmin(L)) and the surface pressures (QA) from which an unloading originates are shown.
Table 1: Leakage data table
The coloured representation corresponds to that in diagram 2. The purple framed fields represent the data obtained from the load branch. These are to be seen in connection with the tightness class (L) next to it and mean that at least the surface pressure must be achieved during assembly (Qmin(L)) in order to comply with the tightness class (L) next to it. The coloured representation should make clear that the values shown in pink represent the intersection points from the loading branch. Below the QA values, which define the surface pressures from which unloading has taken place, the QSmin(L) values are listed for the corresponding unloading branches. These values define the minimum surface pressures in service which must still be present after the gasket has been unloaded in order to maintain the tightness class (L).
Example consideration in connection with diagram 2, with a unloading of 60 MPa: The loading starts at 10 MPa. The next load level is at 20 MPa. Interpolating between the leakage results of both points, the result is an intersection with L = 1E-1 and Qmin(L) = 20 MPa. Further loading from 20 to 40 MPa results in an intersection with L = 1E-2 at Qmin(L) = 33 MPa and then, further loading to 60 MPa results in the intersection with L = 1E-2 at Qmin(L) = 47 MPa. Unloading from 60 MPa to 10 MPa in several steps results in the intersection with L = 1E-3 at QSmin(L) = 12 MPa. These values are transferred to the table.
It is noticeable that the first value (10MPa) in the column of Qmin(L) values is shown in the same colour as the cell frame. There are also several QSmin(L) values in the same color as the surrounding cell frame. This indicates that there is no intersection with the tightness class for this value. However, the value can still be defined. If, for example, the unloading surface pressure in diagram 2 is considered, where unloading starts from 60 MPa, there is the intersection point at L = 1E-3 at 12 MPa. This value is noted in the table at QA = 60 MPa and L = 1E-3. The unloading goes further in this branch down to 10 MPa, but no further tightness class is intersected for this value. In any case, the leakage at 10 MPa maintains L = 1E-2, which in fact includes L = 1E-1, L = 1E-0 and all other tightness classes above. Thus, 10 MPa can be considered the lowest known surface pressure value in this unloading branch and also as the value at which all tightness classes > 1E-3 are maintained.
1: Strictly speaking, at least two tests are carried out for each characteristic value or set of characteristic values for the presentation of the data on the data sheets and mean values are formed and presented. In order not to disturb the reading flow, this fact is not mentioned in the text.
2: All diagrams shown here are fictitious values, which are, however, coordinated in this example.