SUPPLEMENTARY LABORATORY INVESTIGATIONS OF MODERN PLASTIC-POLYMER FISHPLATES FOR RAIL JOINTS

Purpose. The authors’ goal is to determine the behavior of insulated rail joints with polymer-composite fishplates without glueing in the consideration of dynamic loadings regarding to own laboratory tests. In this paper they introduce the applied measurement opportunities. Methodology. Dynamic (fatigue) bending tests were performed by insulated rail joints assembled with plastic-polymer fishplates. The special laboratory measurements are related to digital picture/video measurement technique and assessment method executed by GOM hardware and software, as well as computer tomography according to laboratory bending tests. Findings. In previous papers the authors published the results of glued-insulated rail joints, in this period they continued their research with the investigation of rail joints with plastic-polymer fishplates without glueing. They tested two different types of rail fishplates made of plastic-polymer material. For the rail joints with fishplates but without glueing, the authors applied special measurement techniques by GOM products (Tritop, Aramis) that enable high precision digital measurement techniques with spectacular visualization results. The computer tomography records ensure the opportunity to be able to receive information about inner crackings and faults of plastic-polymer fishplates, with also high precision measurements. The assessment method has to be developed for these specific measurement methodologies to be able to compare the results and define scientific statements. Originality. Up to now any researcher and research group have been dealing with insulated rail joints with special plastic-polymer fishplates without glueing applied mentioned special techniques, no one determined the exact deterioration process of these joints, as well as the crack growing phenomenon in the cross section of the fishplates. Practical value. The research team of the authors had the possibility to see into the details of glass-fibre reinforced resin bonded plastic fishplates during laboratory tests, as well as they publish timely information in the consideration of their laboratory tests’ results. This result can be applied in railway engineering at all stages: design, construction, maintenance&operation in the future.


Purpose
The authors' aim is to define the behavior of insulated rail joints with glass-fibre reinforced plastic fishplates, as well as with and without glue material (between rails and fishplates) regarding to static and dynamic loadings in the consideration of own laboratory tests.
In the authors' previous article [43] the purpose and working mechanism of glued-insulated rail joints were determined. Now the special information related to plasticpolymer material and structures is shortly summarized.
Composite materials or composites are useful materials produced from two or more components with very different physical and chemical characteristics. New material can be made with compound of these parts. Therefore individual characteristics are able to be guaranteed by combination of these components [30]. From other viewpoint: composite material can be given as a combination of a matrix and a reinforcement, which when mixed enable properties better to the properties of the individual parts. In the case of a composite, the reinforcement is the so called fibres and is applied to strenghten the matrix in terms of strength and stiffness [49].
In the aspect of composite materials several structures can be differentiated [33]:  particle-reinforced, o large particle, o dispersion-strengthened,  fibre-reinforced o continuous (aligned), o discontinuous (short),  aligned,  random oriented,  structural o laminates, o sandwich panels. Fibre material can be the followings [19,31]: o glass, o carbon, o aramid, o basalt, o etc. In case of the authors' research the material of railway fishplates is glass-fibre reinforced resinbonded plastic, in that the reinforcement is glassfibre, the matrix is resin.
Two types of glass-fibre reinforced fishplates (fit to 54E1 rail profile) are available for laboratory tests:  type I: structural, laminated polymer ( Fig. 1),  type II: combination of fibre-reinforced polymer with continuous (aligned) and discontinuous, random oriented structure (Fig. 2).
In this paper the authors summarize the up-todate laboratory measurement possibilities and their initial results of plastic-polymer fishplates that are detailed in following sections. Material tests have not been introduced, yet, only in the following publications in 2020.

Methodology
Dynamic (fatigue) bending tests were performed by insulated rail joints assembled with plastic-polymer fishplates. The auhors applied special laboratory measurements that are related to digital picture/video measurement technique and assessment method executed by GOM hardwares and softwares, as well as computer tomography according to laboratory bending tests.
In the following the details of used methodologies and connecting characteristics, parameters are described.
The parameters of investigated fishplates (where type I and II are not specified the data are related to both):   The steps 'iii'…'iv' should be repeated until altogether 3.5 million loading cycles (plan) for 3-3 pieces of rail joints (i.e. 3 specimens with fishplate type I and other 3 with type II).
The results from the initial stage (before fatigue), as well as after each 500,000 loading cycles (after fatigue stages) can be compared together. In this way the crack/failure growing processes are able to be determined and recorded related to the two different fishplates as a function of loading cycles.
In the Findings chapter the authors detail their relevant results.

Findings
In previous papers [24,38,39,40,41,42,43,44] the authors published the results of gluedinsulated rail joints, in this period they continued their research with the investigation of rail joints with plastic-polymer fishplates without glueing.
They tested two different types of rail fishplates made of plastic-polymer material. For the rail joints with fishplates but without glueing, the authors applied special measurement techniques by GOM products (Tritop, Aramis) that enable high precision digital measurement techniques with spectacular visualization results. The computer tomography records ensure the opportunity to be able to receive information about inner crackings національного університету залізничного транспорту, 2019, № 6 (84) and faults of plastic-polymer fishplates, with also high precision measurements. The assessment method has to be developed for these specific measurement methodologies to be able to compare the results and define scientific statements. By this time the following results were obtained with 1-1 pieces of rail joints:  a pre-fatigue tests, i.e. step 'i',  post-fatigue tests, i.e. step 'ii'…'iii', until approximately 10,000 loading cycles.
The reason of only approx. 10,000 loading cycles were applied the fact the fishplates partly or full failured:  fishplate type I: partly failure after 10,000 cycles (one of the fishplate pair is damaged in the middle cross section at the top line),  fishplate type II: failure after 7,331 cycles. Figures 6-7 illustrate the rail joints after 1st loading period.   11 show that the higher the number of (elapsed) loading cycles, the higher the measurable vertical displacement. It is a very trivial behaviour of engineering structures during (and/or after) fatigue test.
The authors demonstrate some of the special measurement results obtained by GOM technology and computer tomography (Fig. 12-15). It can be statedregarding Fig. 12-15 that measurement technique ensured by GOM Aramis is adequate to determine e.g. displacement and strain values very high precision compared to a reference status (so called '0' stage, i.e. nonloaded stage). Every diagram recorded by this method shows the differences. The measurements were executed by 10 Hz sampling while the short dynamic loading was 0.1 Hz. It means that in 10 seconds there was only one full sinus loading cycle, during which 100 pictures were taken. The apparatus of applied, assembled GOM Aramis hardwares and software were able to offer approx. 900 shoots. Because of this fact one measurement took approx. 90 seconds. Fig. 12-15 are typical pictures from the 900 ones. In the future the 'after fatigue' stages have to be recorded to be able to compare the results. (Next to the showed values, the software is able to give not only the vertical, but the horizontal measurements, as well as Epsilon X and Y parametersso called specific strain values.) Figures 16-17 demonstrate some 3-D recordings of computer tomography tests. Referring Fig. 16-17 the authors state that 3-D computer tomography is also adequate for definition inner faults (e.g. crackings, inclusions, etc.) with very high accuracy. The recordings are able to be compared to each other and the deterioration process can be determined by this methodology.

Originality and practical value
Up to now any researcher and research group have been dealing with insulated rail joints with special plastic-polymer fishplates without glueing applied mentioned special techniques, no one determined the exact deterioration process of these joints, as well as the crack growing phenomenon in the cross section of the fishplates. The research team of the authors had the possibility to see into the details of glass-fibre reinforced resin bonded plastic fishplates during laboratory tests, as well as they publish timely information in the consideration of their laboratory tests' results. This result can be applied in railway engineering at all stages: design, construction, maintenance&operation in the future.
Up to now the laboratory measurements with GOM Tritop procedure and the data from those have not been processed yet, but the authors would like to execute it in the future, as well as publish these results.
The authors think that the largest challenge will be the development the data processing and evaluation procedure for both techniques (GOM techniques: Tritop and Aramis, as well as computer tomography).
E.g. in case of GOM Aramis some special points have to be marked before the short dynamic loading during this kind of measurements, after that data have to be filtered/determined from the database and diagrams, figures should be drawn. It means that the change of the behaviour of the fishplated joints can be assessed by usage of the trend functions related to the results from the measurements at different time, i.e. steps (from i) to v), see Section 'Methodology'). These special marked points can be the following on the fishplates:  one point (or more points) from the above zone of the fishplate,  one point (or more points) from the middle zone of the fishplate,  one point (or more points) from the below zone of the fishplate.
The authors have to mention that in case of GOM Aramis the 'points' can be ranges (see Fig. 7 and Figures 12-15). The only requirement to have to be fulfilled: these points or ranges should be able to localized/seen in every recorded picture to be able to define the changes of the parameters related to them.
The second possibility is the GOM Tritop (see Fig. 4). It is a technique with usage of reference points (i.e. without movements/displacements during the measurements), as well as measured points (i.e. they have movements/displacements during the measurements compared to reference points). GOM Tritop gives the opportunity to be able to define the displacement vectors without usage of e.g. Matlab programming. As the authors mentioned, up to now the data processing and assessment have not been performed.
The third possibility is the computer tomography. The recorded 3-D models from computer tomography measurements, the evolution of the crackings or any irregularities inside (or naturally on the surfaces) of the fishplates can be localised and determined. It means that e.g. the length values or maybe the volume (in mm 3 unit) of the faults (i.e. air inside the fishplates), or the number and location of the broken glass fibres are able to be defined. It should be mentioned that computer tomography machine at Széchenyi István University is able to make recordings with limited dimensions, it is the reason the authors focus on the middle part of the fishplates (remark: the highest stress and strain values are in this zone due to the static model and the supports of the 'beam'). The authors have an initial result with this procedure. The volume of the faults (air) related to the Figures 16-17 are the followings:  at the initial stage (i.e. before fatigue test): 3,000 mm3,  after 7,331 loading cycles: approx. 18,000 mm 3 .
It means that the volume of the faults increased the sixfold of the initial after 7,331 loading cycles (after this quantity of loading cycles the fishplates type IIwent broke).
There are some aspects the authors have to consider in the continuation of their research:  specimens should be cut from the fishplates and bending, tensile tests have to be executed (according to the European standards),  from these measurements the material characteristics can be defined,  the performed bending tests with full scale fishplates (see steps in the Methodology chapter) are able to ensure the change of the vertical displacement (deformation) of the rail joints as a function of loading cycles, the elasticity parameters can be calculated (maybe E×I and/or G×A values, Poisson ratio, sigma-epsylonstress-strain -, etc.),  the results will be adequate to compare the behaviour of insulated rail joints with and without glueing, as well as the insulated rail joints with glass-fibre reinforced fishplates and traditional steel fishplates, In the following researchmainly in the preparation of PhD thesis of Attila Némeththe below techniques, methodologies and aspects, have to be considered related to insulated and glued insulated rail joints with glass-fibre reinforced fishplates:  evaluation of geometrical deterioration of ballasted railway tracks [34,35,36,37],  dynamic effects of the railway track (and e.g. turnouts) and vehicles, as well as irregular movements of rail vehicles [3,4,5,6,7,48,50,53],  calculation method of stress-strain rate in the railway layer structures [1,2].