Standard-compliant and low-emission: how GRP flange connections comply with TA Luft

Flange connection components and assumptions 

To take company-specific standards into account, a total of 13 different flange systems were defined for the component test. The selected combinations of collar, loose flange, bolts and gaskets were put together in such a way that they represent the widest possible range of flange systems from real plant operation. 

• Strength design: The strength design of the complete flange connection is part of the TA Luft verification and was carried out or dimensioned for the component test using DIN EN 13121-3 with a safety factor of 6 and a limit elongation for the lining or the chemical protection layer of 0.2 %. For the mathematical verification, the strength verification was carried out based on DIN EN 1591-1. 

• Screws: 25CrMo4 was chosen as the screw material. Washers with a hardness of 200 HV were used for the washers. The transferability of the results to flange connections with bolts of comparable strength, such as stainless steel A2-70, is given. For bolts with lower strength, an individual transferability test is required. 

• Loose flanges and collars: Previous investigations have shown that loose flanges made of GRP materials do not provide the required strength to achieve the necessary surface pressures for effective sealing. Therefore, metallic loose flanges were used for the tests. 

The collars used in the test series were provided by Kurotec and are based on the dimensions of the current draft revision of DIN 16965. Although this standard also provides for collar geometries with reduced blade thickness for elastomer gaskets, only variants with reinforced blade thickness were used in the tests carried out. 

As standardized standard components were used in the project, the TA-Luft verifications provided do not relate to a specific GRP manufacturer. 

• Lining material: Material groups were formed when selecting the lining material for the collars. For example, tests with a PE-el lining based on PE 80 can be transferred to PE 100. 

• Seals: The study considered two groups of gaskets: Elastomer profile gaskets (EPDM) with a steel insert and expanded PTFE flat gaskets (ePTFE). The selection of these gasket groups is based on the factory standards of the chemical companies involved and the specific gasket characteristics according to DIN EN 13555. This corresponds to the common gasket types used in the industry for flange connections in GRP pipe systems. 

• Test temperature: The selection of the exposure temperatures for the component test was based on the maximum application temperature in accordance with DIN 16867:1982 Table 1. For operating conditions below 85 °C, the load-bearing laminate Derakane 411-350 was used; for temperatures above 85 °C, the load-bearing laminate Derakane 470-300 was used. 

• Test pressure: The test pressure of 16 bar selected for the test series is based on Table 1 of DIN 16867:1982. 

Table 1 summarizes the lining materials and the associated operating temperatures and test pressure. 

Verification of TA Luft via type-based component tests 

The original scope of the project was to verify the tightness in accordance with the requirements of TA Luft by means of component tests - based on VDI 2290 and the type-based test method described by the VCI. 

The tightening torques for screws when using elastomer or PTFE seals were determined in accordance with DIN EN 13121-3 (2016 edition). The design assumed of an installation surface pressure of 10 MPa and an operating surface pressure of 5 MPa for PTFE gaskets. The torques determined can be found in Table 2. 

Carrying out the component tests 

As part of the component test, the assembled DN40 flange connection is first filled with helium, the test pressure is set to 16 bar and an initial leakage measurement is carried out at room temperature. A special measuring cartridge is used to minimize interfering influences due to permeation of the lining and the GRP material. The leakage rate is recorded continuously over a period of 24 hours using a helium mass spectrometer in vacuum mode. 

The flange connection is then heated at ambient pressure to the intended ageing temperature, which is kept constant for 48 hours. After cooling down to room temperature, another 24-hour leakage measurement is carried out under the specified test pressure, again using the helium mass spectrometer. A limit value of 0.01 mg/(s-m) applies to all measurements to comply with the TA-Luft specifications. 

Results of the leak tests 

In the tests with EPDM gaskets, the tightness criterion was always met after the flange connection was installed in accordance with the installation specifications in Table 2. For the material combination GRP with chemical protective coating (CSS), the test with a GRP loose flange was also successful. It was not necessary to retighten the flange connections during the test. 

The situation was different in the tests in which an ePTFE gasket was used. In some cases, the connections were already so leaky during the first filling that the test had to be aborted. In other tests, the leakage rates were no longer permissible after the temperature exposure. Tests were also carried out with reduced ageing temperatures (60 °C) for the PP/GFRP and PE-el/GFRP composites. In the case of PP/GRP, the temperature reduction was sufficient to pass the TA Luft test without re-drawing. For PE-el/GFRP, the test at 60 °C was not successful. 

All results of test series 1 are summarized in Figure 3. 

Verification of TA Luft via mathematical method 

After the failed component tests with PTFE gaskets, calculations were carried out in accordance with DIN EN 1591-1 to check the torques in Table 2. The calculation method of DIN EN 1591-1 can be applied to flange connections with round flanges, bolts and gaskets; consideration of non-metallic materials is not explicitly excluded. With the minimum requirements for the strength of GRP components described in DIN 16965 / DIN EN 13121-6:2016, strength parameters are now available for the first time that can also be used in the calculation method of DIN EN 1591-1. For this purpose, the safety coefficients defined in DIN EN 13121-3 for the GRP collars were applied in the calculation routine. The behavior of the lining materials is not considered in the calculations, but it would be conceivable to take the different creep/relaxation properties into account. 

The DN40 flange connection served as a reference for determining the tightening torques in test series 2. The calculated tightening torque of 100 Nm generates an installation surface pressure of around 55 MPa. To be able to transfer these results to other nominal sizes, it was necessary to ensure that at least comparable surface pressures were applied to the gasket during installation. 

The tightening torques derived from this for different nominal sizes were adapted as far as possible so that uniform assembly specifications could be achieved for identical bolt sizes. However, complete standardization was not possible for bolt sizes M16 and M24, as it was not possible to define a common tightening torque that both met the requirements for tightness and was below the maximum permissible value - as illustrated in Figure 4. 

Verification of calculations by means of type-based component tests 

The outstanding component tests with PTFE gaskets were then carried out with the newly determined torques. The purely mathematical, theoretical proofs were thus verified experimentally by means of component tests. 

As the overview in Figure 5 shows, the sealing behavior of the flange connections was improved by increasing the tightening torque. However, the tightening torque of 46 Nm, which emerged as the lower limit value from the mathematical design, was not yet sufficient to meet the requirements of TA Luft. At a tightening torque of 100 Nm, the limit value of 0.01 mg/(s-m) was finally complied with for all flange material combinations with ePTFE gaskets. A lower tightening torque might have been sufficient, but to minimize the number of tests, no further intermediate steps were tested. It was not necessary to retighten the flange connection here either. 

Summary 

To meet the tightness requirements of TA Luft for GRP flange connections, the flange geometries were adapted regarding their sheet thicknesses and the corresponding strength parameters were redefined within the applicable standards. On this basis, it was demonstrated that both a type-based component test and a mathematical verification in accordance with DIN EN 1591-1 can be used as suitable methods for assessing tightness. 

However, the component test, in which the temperature influence is simulated by a single exposure in the oven, only allows limited statements to be made about the long-term behavior of the components under real operating conditions. Especially in applications with frequent temperature fluctuations, it may be necessary to retighten the flange connections. The operators involved in the project will adapt their individual assembly specifications based on the newly determined tightening torques and test the implementation in operation. 

Even if not all GRP flange systems described in the regulations were included in the study, the results obtained provide valuable findings for the construction and design of GRP flange connections that meet the stricter emission requirements of the revised TA Luft.