1 Introduction

Ubiquitous industries (e.g., chemical, petrochemical, the food and pharmaceutical sectors) use various types of heat exchangers (HE) to assure thermal control of their production lines [1, 2]. Among the available models, the gasketed plate heat exchanger (GPHE) is frequently chosen owing to its compactness and flexibility concerning design possibilities. The GPHE structure consists of plates (through which the heat exchange is carried out), gaskets, thick compression end-plates and bolts for tightening. The sealing mechanism of these systems guarantees the sealing of processing fluids under extreme environments. While subsea oil and gas production uses metal-to-metal sealing mechanism due to extreme environmental conditions [3], elastomeric gaskets provide sealing in GPHE systems [1].

Despite the use of metallic plate materials such as stainless steel and titanium which allow high temperature operations, gasket materials used in GPHE determine maximum working temperatures in these systems. Gaskets can be affected by environments containing oxygen, which roots rubber oxidation [2, 4, 5], and allow operating temperatures up to 200 °C for some specific gasket materials. According to Adolfsson [5], breakdown time along GPHE useful life is nearly 30%, whereas 68% of failures are related to gasket expulsion or the lack of gasket sealing capacity due to degradation of their properties, which highlights the importance of these components [6, 7]. Recent studies about GPHEs address heat exchanger failure modes and progress in these structures [5, 8], like corrosion in stainless-steel plates [9, 10], fatigue [11, 12]. Other studies with GPHE focus on thermal and hydraulic performance under operation conditions, as well as the effects of plates geometry and operation parameters [13, 14]. Rubber materials studies have been recently studied in order to predict the O-rings and gaskets lifetime [15, 16] based on accelerated thermo-oxidative aging [17,18,19,20].

Common gasket materials used in oil and gas industries are nitrile butadiene rubber (NBR), hydrogenated nitrile butadiene rubber (HNBR), ethylene-propylene-diene rubber (EPDM), fluoroelastomer (FKM) e perfluoroelastomer (FFKM) [21]. NBR and HNBR materials are characterized by elevated chemical resistance, allowing applications in hazardous environments [22, 23]. EPDM rubbers are not used in applications involving oil; however, they support higher temperatures than other typical gaskets [24, 25]. FKM are commonly studied for use in extreme environments, like high temperatures [26].

The sealing capacity of these rubbers is seldom investigated in heat exchangers applications. Some recent researches address the sealing performance for hydrogen industry [27], including focusing on the composition of rubbers [28]. Applications for deep sea and high temperature for nitrile rubber also have been studied [29, 30]. Kömmling et al. monitored the variation of HNBR and EPDM properties for 1.5 years and related gasket failures with intense material property changes [16]. Liu and Lian investigated rubber failures for sealing mandrel hanger. They identified the reduction in rubber elasticity by monitoring internal stresses for weights above 150 ton [31]. Zheng and Li studied the behavior of HNBR rubbers by means of stress relaxation tests and finite element analysis [32]. Zuo et al. compared the sealing performance based on contact stress for HNBR sealing components by employing theoretical model and the finite element model considering the stress relaxation effect are established [33]. Hu et al. studied the sealing capacity packaging element and shown the existent relationship between axial load due to packing compression and the maximum difference pressure [34].

This work assesses the sealing behavior of HNBR and EPDM gaskets applied in GPHEs. Compression strength experiments with the aid of a prototype reproduce the GPHE assembling procedure. GPHE integrity and mechanical characteristics were evaluated with stain gauge measurements in HE plates and sealing experiments. The effects of tightening distance and sealing pressure on gasket performance were determined.

2 Experimental procedure

2.1 Materials

Experiments were executed with gasket plate heat exchangers consisting of HNBR and EPDM sealing materials, and stainless steel plates (SAE 316L). The corrugated wavy pattern plate shape is an outcome of cold forming. Martins et al. characterized the applied 316 plates using tensile tests on standardized tensile specimens extracted from the plate region where the gasket is present according to ASTM Standard Test Methods [35, 36]. Young's modulus (E) and Poisson's ratio (ν) were determined as 128 ± 12 GPa, and 0.31, respectively [35].

Gasket performance evaluation occurred with strain gauge measurements in different plate locations during GPHE assembly and with sealing tests, and with compression strength experiments in prototypes formed by gasket and plate segments. In the former, entire gaskets accommodated in the original plates were assessed with varying tightening distances and sealing pressures (Fig. 1a). In the latter, continuous circular gasket segments extracted from the branches of the GPHE inlet/outlet were coupled to corresponding plate segments (Fig. 1b).

Fig. 1
figure 1

Photographs of a GPHE plate and gasket and b gasket and plate segments

Full size image

Gasket material properties are defined as follows. Nitrile-based material (HNBR) comprises 96% hydrogenation, 36% acrylonitrile (C≡N), and an iodine value of 11 phr, while the EPDM rubber consists of 55 phr ethylene and 2.3 phr ethylidene norbornene (ENB). Gasket heights of both materials were determined as 4.113 ± 0.110 mm with a digital micrometer from Mitutoyo DSC250-X. Hardness values of 78.1 ± 0.6 and 81.5 ± 0.5 Shore A for HNBR and EPDM rubbers, respectively, were obtained with Mitutoyo HH 336 Shore A durometer with a standard measuring support (Mitutoyo 811–013), following ISO 7619–1 standards [37]. The gaskets have a complex cross-sectional geometry with nominal height of 4 mm and nominal width of 9 mm [15]. The prototype has an internal diameter of 110 mm which corresponds to 14% of the total length of a full-scale joint for this GPHE model.

2.2 Assembly

The assembly process of plate heat exchangers follows progressive compression stages so that gaskets reach the operating compression level determined by the design tightening distance. Table 1 presents typical compression levels obtained during the GPHE assembling procedure with the actual model of gasket and plates. The level 1.00A represents the ideal tightening distance (maximum tightness) specified by the manufacturer.

Table 1 Typical compression levels obtained during the GPHE assembling procedure
Full size table

The assembly procedure follows Martins et al. [35]. Once a tightening distance is applied over the heat exchanger, periods of ten minutes allow gasket stress relaxation. The succeeding reduction in gasket stresses reduces the necessary work to reach the operating compression levels. Once the assembly is concluded and before the sealing tests are initiated, a thirty-minute interval provides the final relaxation.

2.3 Compression strength

The prototype was subjected to compression stress relaxation experiments by means of Instron 23–100 universal testing machine equipped with a 100 kN load cell (Instron CCE 100kN). The compression strength generated during the assembly was monitored. Porthole segments extracted from real-scale gaskets characterize the prototype specimens (Fig. 1). The continuous porthole gasket segment was selected since it consists of a closed boundary element and it can fit in the space designated for compression loads. Therefore, this element can better mimic the behavior of the closed real-scale gasket (Fig. 2).

Fig. 2
figure 2

Photograph of the prototype arrangement subjected to compression loads

Full size image

The test temperature was set to 30 °C and controlled by a thermal chamber with an accuracy of ± 1 °C. Changes on the tightening distance (in accordance to Table 1) occurred with a displacement rate of 1 mm.min−1. Following ISO 3384, the measured strength in relaxation experiments consists of values recorded 30 min after the desired compression plateau is reached [39]. Therefore, the measured stress after relaxation is the compression strength regarding the complex geometry of the prototype. Sets containing 1–16 gaskets provide the effect of the number of gaskets on the compression strength. With the above procedure, it is possible to identify the number of gaskets which provide a representative condition for GPHEs’ behavior.

2.3.1 Stress analysis

Deformation measurements at specific plate locations were obtained with electrical resistance triaxial strain gauges during GPHE assembly. Stress is then calculated in the elastic regime as applied by Martins et al. [25]. Strain gauge positioning followed the works of Martins et al. and Nascimento who presented GPHE main failure mechanisms [38, 40]. The strain gauges (SG) were positioned at the heat exchange region (SG1), the fluid distribution area (SG2) and the gasket groove of the porthole region (SG3), as shown in Fig. 3. Monitoring of plate deformations occurred in assembly experiments with EPDM and HNBR gaskets.

Fig. 3
figure 3

Positioning of triaxial strain gauges at the heat exchange region (SG1), the fluid distribution area (SG2) and the gasket groove of the porthole region (SG3)

Full size image

In the elastic region, stresses (σ) increase with increasing strain (ε) according to Hooke’s Law. Due to the plate complex geometry, the principal strain directions are unknown. By considering isotropic material, the principal stresses (σp,q) are calculated as:

$$\sigma_{p,q} = \frac{E}{2}\left[ {\frac{{\varepsilon_{0} + \varepsilon_{90} }}{1 - \nu } \pm \frac{{\sqrt 2 { }}}{1 + \nu }\sqrt {(\varepsilon_{0} - \varepsilon_{45} )^{2} + (\varepsilon_{45} - \varepsilon_{90} )^{2} } } \right]$$
(1)

where the subscripts 0, 45 and 90 stands for strain directions. Multiaxial von Mises criterion determines von Mises equivalent stress (σVM) as:

$${\upsigma }_{{{\text{VM}}}} = \sqrt {\frac{{{\upsigma }_{p}^{2} + {\upsigma }_{q}^{2} + \left( {{\upsigma }_{p} + {\upsigma }_{q} } \right)^{2} }}{2}}$$
(2)

2.3.2 Equivalent stiffness

The unidirectional load applied in the prototype acts in the normal direction to the plate surface as illustrated in Fig. 2. Therefore, estimation of the gasket stiffness in the load direction is a simplification for obtaining the gasket compressive strength response and GPHE’s stiffness, see Kelly e Konstantinidis [41].

The vertical stiffness of the elastomeric structure (\(K_{g}\)) consists of combining N gaskets in series. The stiffness determination is given by, Van Engelen [42]:

$$K_{g} = \frac{{E_{c} A_{{{\text{cs}}}} }}{{y_{0} }}$$
(3)

where \(A_{{{\text{cs}}}}\) is the vertical cross-sectional area, \(y_{0}\) is the gasket height and \(E_{c}\) is the rubber compression modulus. The latter is a parameter related to the rubber shear modulus (G) and the shape factor (S):

$$E_{c} = 4GS^{2}$$
(4)

For rubbers without holes, \(E_{c}\) is equal to \(6GS^{2}\). However, considering ratios between inner and outer diameters greater than 0.5, the hole influence is significant and modifies Ec values to \(4GS^{2}\), see reference [41]. The rubber shape factor (S) for annular geometry considers the ratio between the loaded area (cross-sectional area) and the free-surface area:

$$S = \frac{{\left( {d_{o} - d_{i} } \right)}}{2t}$$
(5)

where \(t\) is the rubber thickness, and \(d_{i}\) and \(d_{o}\) are the inner and outer diameters, respectively. By employing Eq. (4), different shear moduli result from different compression levels and number of gaskets. Note that \(E_{c}\) is related to rubber shear modulus and shape factor; however, it does not represent gaskets properties but the combination of rubbers and plates in series, see reference [43].

The equivalent compression modulus (\(E_{{{\text{eq}}}}\)) can be calculated as a function of rubber compression and cross-sectional area \(\left( {A_{{{\text{cs}}}} } \right)\) [41]:

$$E_{{{\text{eq}}}} = \frac{P}{{A_{{{\text{cs}}}} \varepsilon_{y} }}$$
(6)

where P is the instant applied load and \(\varepsilon_{y}\) is the instant rubber vertical strain, obtained by:

$$\varepsilon_{y} = \frac{{y_{0} - y}}{{y_{0} }}$$
(7)

where y is the instant height. By recording the load P 30 min after reaching each compression level mitigates viscous dissipation effects during rubber stress relaxation.

Despite the approach simplicity, it allows sealing capacity relationship between prototype and real-scale heat exchanger. Evaluation of heat exchanger sealing capacity is costly and time-consuming.

2.4 Sealing tests

A critical condition in experiments with GPHEs considers only one pressurized branch (i.e., single-test condition), while another is free for deformation, which promotes elevated mechanical stresses [35, 38]. Regarding the pressurization systems, GPHE tests occurred with Flutrol equipment consisting of a hydro-pneumatic pump type Haskel ASF-35 (Fig. 4a), while pressurization of the prototype occurred with a Flutrol TestPac 33, consisting of a Haskel MS-71 hydro-pneumatic pump (Fig. 4b). The prototype comprises a pack of 6 gaskets. Hydrostatic experiments were performed only after fulfilling the sealing structures with water in ambient temperature (T ~ 23 °C), ensuring the air removal. An acquisition system and a pressure transducer (Omega PX409 750 psig) monitored the instant pressure from the inlet branch, controlled by a manual valve.

Fig. 4
figure 4

Pressurization test benches for a GPHE [38] and b prototype [15]

Full size image

The pressure increment was adjusted to two bars. Three-minute interval was sufficient to detect leakage in both systems before reaching the next pressure steps, which characterizes an overall pressurization rate of 0.66 bar.min−1. Static pressure reduction of 10% or more characterized a leakage event [15]. The pressure increase occurred for the same compression level. Experiments occurred with different tightening distances, ranging from 1.30 to 1A.

Experiments with GPHEs comprise a twenty-plate pack with the same gasket material. GPHEs can operate without leaks within a wide range of tightening distances stipulated by the manufacturer. The minimum tightening distance represents the leak-tight distance. However, during the operation of GPHEs, there are chances of leak events between minimum and maximum distances due to gasket aging or gasket displacement [15, 19].

3 Results and discussion

In this section, all experiments were evaluated with the aid of analysis of variance (ANOVA) within a confidence interval of 95%. Each experimental point is an average of five repetitions. Maximum standard deviation regarding the whole set of experiments is 5.1%.

3.1 Compression strength

Figure 5 presents the results of compression strength owing to stress relaxation experiments as function of the number of plates (N) with different tightening distances.

Fig. 5
figure 5

Effects of tightening distance and number of gaskets on the HE compression strength: a EPDM and b HNBR gaskets

Full size image

Compression strength increases with decreasing tightening distance, since the rubber strain also increases. The compression strength also increases by increasing the number of gaskets (N), when N ranges from 1 to 6. However, for prototypes consisting of 6 gaskets or more, a rigid body behavior stabilizes the pack’s compression strength due to stiffness reduction [44, 45], as observed in Fig. 6. The compression strength is also affected by the interaction between the elastic (plate) and viscoelastic (rubber) components. HNBR compression strength reaches nearly 14 kN with experiments consisting of 6 gaskets or more, while EPDM gaskets reach about 30 kN.

Fig. 6
figure 6

Combined effects of compression strain and number of gaskets on the HE stiffness for EPDM and HNBR gaskets

Full size image

Figure 6 shows the effects of compression strain and number of gaskets on the heat exchanger stiffness for EPDM and HNBR gaskets.

The GPHE stiffness increases with increasing compression strain and decreasing number of gaskets. EPDM gaskets show superior compression resistance and stiffness values as compared to those obtained with HNBR gaskets. This outcome is also related to superior hardness values for EPDM material as informed in Sect. 2.1, which results in a superior restriction to the mobility of rubber chains [46, 47].

Figure 7 shows the effect of compression strain on compression strength (P) for EPDM and HNBR gaskets (left y-axis). The dashed line stands for PEPDM/PHNBR strength ratio (right y-axis). Images at the top qualitatively represent the decreasing tightening distance (regarding tightening steps) from left to right and the adaptation of the rubber in the groove.

Fig. 7
figure 7

Effect of compression strain on compression strength (P) for EPDM and HNBR gaskets (left y-axis). The dashed line stands for PEPDM/PHNBR strength ratio (right y-axis)

Full size image

Compression strength increases with increasing compression strain (see black and red lines) and decreasing tightening distances (see images at the top). Since the physical restriction imposed by the plates to gaskets’ deformation increases with decreasing tightening distance, the compression strength increase tends to be exponential. When the tightening distance ranges from 1.3 to 1.1A (or compression strain ranges from 0.17 to 0.27), the gasket material has elevated influence on compression resistance, i.e., the compression modulus approaches the properties of the material (rubber) since the aspect ratio is relatively low [42]. When the tightening distance ranges from 1.1 to 1.0A (or compression strain ranges from 0.27 to 0.33), the rubber influence on the final compression resistance decreases, i.e., the gasket influence on the compression modulus (\(E_{c}\)) depends on the material, the restriction to deformation, and the thickness reduction, which affect the shape factor. Note that the strength ratio, PEPDM/PHNBR (right y-axis), decreases for compression strain over 0.27: gasket properties become less important.

In summary, the rubber compression strength during the tightening procedure is an outcome of the combined effects of the stiffness variation due to the number of gaskets, tightening distance, gasket material properties, and restricted deformation.

3.2 GPHE assembling

Table 2 summarizes the stress levels obtained for five tightening distances in three different plate locations: heat exchange area, flow distribution area and nozzle, for EPDM and HNBR gaskets. Stress levels remained constant during the relaxation periods, as shown by references [35, 38]. When the tightening distance is 1A, the gasket height is equivalent to 2.6 mm.

Table 2 Effects of tightening distance and plate local area on Von Mises stress levels during heat exchanger assembly for EPDM and HNBR gaskets
Full size table

The nozzle region has staggered gaskets to allow the inlet and outlet of working fluids. These gasket segments exert significant contact pressure on the plate to ensure sealing capacity, increasing stress levels. Consequently, the gasket type and material significantly influence the generated stress. Highest stress levels were obtained with EPDM gaskets due to elevated values of hardness, stiffness and compression strength. In the nozzle region and with the tightening distance equal to 1A, plate stress level was equal to 316 MPa for EPDM gasket and to 133 MPa for HNBR material. Superior stress levels for EPDM application on GPHE (2.38 times higher than HNBR application) agree with compression strength values previously presented (2.14 times higher for EPDM material as compared to HNBR one).

Figure 8 shows the effect of the tightening distance on GPHE von Mises stresses (left y-axis) and on the prototype compression strength (right y-axis). The results of GPHE stress levels are presented at the most requested region (nozzle area). The compression strength results are shown for prototypes with six gaskets.

Fig. 8
figure 8

Effect of the tightening distance on GPHE von Mises stresses (left y-axis) and on the prototype compression strength (right y-axis)

Full size image

Von Mises stresses and compression strength increase with decreasing tightening distance. Note that the compressive strength provided by EPDM gaskets yields substantial stresses as compared to HNBR gaskets.

3.3 Sealing tests

Figure 9 shows the effect of compression strain on prototype and GPHE sealing capacity (or maximum supported pressure before leakage occurrence) for EPDM and HNBR gaskets. Black lines are related to EPDM results, and red lines, to HNBR results. The results are shown for prototypes with six gaskets and GPHE with 20 gaskets.

Fig. 9
figure 9

Effect of compression strain on prototype and GPHE sealing capacity (or maximum work pressure) for EPDM and HNBR gaskets

Full size image

Leakage occurs in the interface between the plate and the gasket. The sealing capacity increases with increasing compression strain. High compression strains result in high contact stresses and compression strength, yielding superior sealing capacities and supported pressures. With the same compression strain, higher working pressures can be attained with EPDM gaskets owing to higher compression strengths as compared to HNBR gaskets. Furthermore, with the same compression strain, supported pressures in prototypes are more significant than the ones obtained with GPHEs. The structure that embraces the prototype is compacter and stiffer than the GPHE structure. Note that the prototype comprises a uniform and continuous geometry (circular) with greater robustness, while the GPHE has complex contact stress distribution along the gasket. The GPHE frame has been designed for working pressures up to 20 bar, and, therefore, it was not possible to identify the sealing capacity of GPHE experiments with EPDM gaskets when the compression strain was equal to 37% (corresponding to the tightening distance of 1A).

Table 3 shows empirical correlations obtained by linear regressions for sealing capacities (or maximum working pressures) of HNBR and EPDM gaskets, regarding experiments with prototypes and heat exchangers. The variables P, G and C denote prototype maximum pressure, GPHE maximum pressure and compression strain, in that order. The subscripts “E” and “H” stand for EPDM and HNBR gaskets, respectively. Equations that relate EPDM/HNBR and prototype/GPHE results are also provided. The quality of the correlations is expressed by R2-values over 0.94. From these relationships, it is possible to estimate the maximum operating pressure of the heat exchanger based on tests performed on prototypes considering the gasket material used.

Table 3 Sealing capacities (or maximum work pressures) for HNBR and EPDM gaskets
Full size table

4 Conclusions

This study investigates the structural performance and sealing capacity of gasketed plate heat exchangers (GPHE) equipped with hydrogenated nitrile butadiene rubber (HNBR) and Ethylene Propylene Diene Monomer (EPDM) gaskets. Appraisal occurs with sealing and compression strength experiments, and strain gauge measurements in heat exchangers and prototypes representing these structures.

The rubber compressive strength during the tightening procedure is an outcome of the combined effects of the stiffness variation due to the number of gaskets, tightening distance, gasket material properties, and restricted deformation. Compression strength increases with increasing compression strain and increasing number of gaskets. However, experiments in prototypes revealed stable strength levels for tests with six plates or more.

Regarding the employed gasket materials, prototypes with EPDM gaskets showed 30 kN in the last tightening step (1A) of the compressive strength tests compared to 14 kN with HNBR ones. This outcome is also related to superior hardness values for EPDM material, which results in a superior restriction to the mobility of rubber chains. The resulting stiffness values ranged from 0.3 to 7.0 kN/mm.

Comparing to heat exchangers experiments, the prototype compression strength is directly related to the elevated stress levels measured at specific plate areas during tightening procedure. During tightening procedure with the real-scale heat exchanger, the resulting von Mises stress level in critical region was 316 MPa and 133 MPa using EPDM and HNBR gaskets, respectively. Since von Mises stresses increases with decreasing tightening distance, we conclude that the heat exchanger stiffness increases with compression strain and decreases with number of gaskets.

The sealing capacity of the evaluated systems increases with increasing compression strain, as well as compressive strength and supported working pressures. Therefore, with the same compression strain level, higher working pressures can be attained with EPDM gaskets owing to higher compressive strengths as compared to HNBR gaskets. Finally, empirical correlations were obtained to relate sealing capacity and compression level for EPDM and HNBR gaskets. These correlations allow the relationship of the sealing capacity between prototype and real-scale heat exchanger, since the sealing capacity assessment of the heat exchanger is costly and time-consuming.