Advertisement

Lightweight Design worldwide

, Volume 10, Issue 4, pp 6–11 | Cite as

Investigation of interactions of CFRP with processing fluids

  • Anna Lena Demmerling
  • Hans Jürgen Schlindwein
  • Sebastian Geier
  • Thorsten Mahrholz
Cover Story Machining of CFRP
  • 184 Downloads

In the machining lightweight sector prejudices against influence of processing fluids to material properties of FRP have still prevailed. Rhenus Lub, developer and producer of lubricants, has investigated more detailed the interactions between special processing fluids for composites and CFRP in cooperation with the German Aerospace Centre.

In metalworking industry the application of metalworking fluids during machining processes is widely spread. In cooling and lubricating the contact zone between tool and workpiece not only an extension of tool life but also an improvement of surface qualities is possible [1]. For fibre reinforced plastics (FRP) processing fluids are only rarely used. According to a survey of the year 2014 more than 50 % of the users machine dryly and only 16 % use flood cooling [2].

Because of very small and respirable particles encapsulation and extraction are necessary during dry machining [3]. Delamination and protruding fibres are frequent results of dry machining. Tool wear is high due to high abrasiveness of the hard fibres so that high-priced tools made of polycrystalline diamond or with special coatings have to be used.

Benefits of Wet Machining

In case of FRP the benefits of wet machining are various. Research projects showed that dust was bound and tool wear decreased by application of processing fluids [4]. Surfaces of machined CFRP workpiece edges could be improved significantly using specially developed processing fluids in machining tests [5]. During an industrial field test of machining composites the feed rate could be raised by 60 % only by using a processing fluid [6].

Interactions between processing fluids and CFRP have been rarely investigated. From investigations concerning the interlaminar shear strength resulted that this property doesn’t change significantly during the impact of an 8 % processing fluid for 96 h [4]. Further material relevant properties concerning the connection of fibre and matrix can be tested during a transverse tensile test. The German Aerospace Centre (DLR) conducted corresponding experiments using processing fluids specially developed by Rhenus Lub.

Surfaces of machined CFRP workpiece edges could be improved significantly using specially developed processing fluids.

Processing Fluids to Machine Composites

At the DLR two different processing fluids (PF) newly developed for the machining of FRP were tested. The PF rhenus XT 46 FC is water mixable, contains mineral oil, and builds a very finely dispersed oil-in-water emulsion whereas rhenus XY 190 FC is a full synthetic PF clearly solving in water. The suggested application concentrations are 8 %. In industrial field tests the PF rhenus XT 46 FC succeeded even in machining of aluminium and titanium. Therefore it can be very well applied for the machining of stacks. For the tests one part PF concentrate was mixed with nine parts water.

Material Testing as “Worst-case” Scenario

Currently there are a lot of reservations about widespread use of PF in the composite working sector. These are caused by the fact that reactions with the resin cannot be excluded generally whereas the fibres themselves are inert. To investigate interactions between PF and CFRP the DLR prepared material samples made of epoxy based CFRP. Plates of unidirectional carbon fibre fabrics (Style 796; Firma ECC Cramer) and a three components epoxy resin (Araldite LY 556, Aradure HY 917, DY 070; Firma Huntsman) were produced by injection process (DP-RTM [7]).

Geometry and homogeneity of the plates were examined by ultrasonic technique. The fibre volume ratio was about 60 %. The sample dimensions according to DIN EN 2561 were produced by milling the surfaces and trimming the edges. Matrix dominating transverse tensile samples were produced to investigate very sensitively the effect of PF.

Reference samples were machined dryly and test samples were machined by applying PF. Subsequently the test samples were stored at 60 °C for one week in 10 % PF. Because water evaporates during the PF’s application the concentration was chosen higher than the proposed application concentration. Thus the chosen test conditions mean a “worst-case”-scenario because contact times in practice are essentially less and machining of one workpiece rarely lasts longer than 8 h. After storing the samples were cleaned in an ultrasonic bath filled with demineralised water for 30 s. A subsequent bath in butanone for 60 s ensured surfaces free of grease to glue on cap strips.

Transverse Tensile Tests as Investigation of Fibre-Matrix-Connection

According to DIN EN 2561 tensile force is initiated orthogonally to the fibre direction and is increased continuously until sample breakage. Thus not the fibre’s stiffness but the cohesion of fibre and resin is tested. The geometry of a material sample is schematically depicted in Figure 1.
FIGURE 1

Geometry of a transverse tensile sample according to DIN EN 2561 (© Rhenus Lub)

Using the measured values force F and elongation ε and the known cross-sectional area S, strength σ, and stiffness E can be calculated (according to [8]):
The tests were conducted in a Zwick testing machine (type 1476 und 1484) at room temperature and − 50 °C. The averaged property values each based on minimum six single measurements are plotted in Figure 2.
FIGURE 2

Material properties of conditioned transverse tensile samples (storage in 10 % PF, 168 h, 60 °C) (© Rhenus Lub)

From the bar chart can be derived that the material properties of the test samples treated with PF differ from those of the dryly treated reference sample. Generally the material properties of the test samples are below the material properties of the reference sample. To judge if these differences are characteristic or statistically random the significance is analysed with a probability of failure of 1 % (α = 0.01) [9]. A two-tailed Student t test was conducted assuming same variances. That level (p-value) was calculated that shows the average values just now being significantly different. If the calculated p-values lie above the probability of failure, the checked average values do not differ significantly. The determined differences are statistically random. Results and conclusions of individual tests are summed up in Table 1.
Table 1

Significance analysis of the measured values at a probability of failure α = 0.01 (© Rhenus Lub)

Material property

Temperature

PF (10 %)

p-value

Difference between mean values is …

Ultimate elongation

23 °C

rhenus XT 46 FC

0.5913

not significant

rhenus XY 190 FC

0.1175

not significant

−50 °C

rhenus XT 46 FC

0.9918

not significant

rhenus XY 190 FC

0.1909

not significant

Transverse tensile strength

23 °C

rhenus XT 46 FC

0.0729

not significant

rhenus XY 190 FC

0.283

not significant

−50 °C

rhenus XT 46 FC

0.6965

not significant

rhenus XY 190 FC

0.0263

not significant

Transverse tensile stiffness

23 °C

rhenus XT 46 FC

0.6251

not significant

rhenus XY 190 FC

0.0004

significant (−2.4 %)

−50 °C

rhenus XT 46 FC

0.2780

not significant

rhenus XY 190 FC

0.1962

not significant

Currently there are a lot of reservations about widespread use of PF in the composite working sector.

The test samples stored in rhenus XT 46 FC did not show any irregularity at 23 °C as well as −50 °C. In consideration of a 99 % confidence probability the deviations of the determined average values are not significant. Looking at the values concerning rhenus XY 190 FC the material properties at −50 °C do not differ from those of the reference sample. For rhenus XY 190 FC the transverse tensile stiffness decreases at 23 °C by 2.4 %. Ultimate elongation and transverse tensile strength do not indicate significant differences at 23 °C. Thus rhenus XT 46 FC provides slightly better results in the transverse tensile test than rhenus XY 190 FC. As a possible cause for the decrease of stiffness at 23 °C the favoured diffusion of rhenus XY 190 FC into the composite was discussed. A fluorescence marker was added to the PF before the storage and the samples were investigated by fluorescence microscopy with 10-fold enlargement. An evidence for a penetration of the PF could not be provided.

In view of the extreme testing conditions (10 % PF, 168 h storage, 60 °C) can be assumed that any significant deviations concerning transverse tensile strength, stiffness, or ultimate elongation could not be determined using these PF in real machining applications. A wet machining of epoxy-based CFRP with rhenus XT 46 FC and rhenus XY 190 FC can be judged as uncritical.

Because many workpieces are sealed after the machining process, a very good wash-down of the PF is fundamental.

Wash-down of processing fluids

Because many workpieces made of CFRP or GFRP are sealed after the machining process, a very good wash-down of the PF is fundamental. The efficacy of individual washes was tested by the DLR. After storage at 60 °C for one week the samples were cleaned in consecutive steps. After each cleaning step the samples were dried in a drying cabinet at 80 °C.

The samples of the first test series were washed with demineralised water. The samples of the second test series were washed with demineralised water and were cleaned in an ultrasonic bath filled with demineralised water for 30 s. For the third test series the samples were additionally immersed in a solvent bath filled with butanone for 60 s, Figure 3. The relative sample weight after each cleaning step is presented in Figure 4.
FIGURE 3

Cleaning steps after storage (© Rhenus Lub)

FIGURE 4

Effectiveness of several cleaning steps, cleaning with water is sufficient (© Rhenus Lub)

For a suitable plotting the sample weights were scaled to the respective starting weight. Attention should be paid to the very small range of the ordinate of 0.006 or 0.6 %. After storage a low increase of the starting weight can be recognised. The weight of samples stored in rhenus XY 190 FC increases by 0.05 %. For rhenus XT 46 FC it is about 0.11 %.

After a wash-down with demineralised water the sample weight of rhenus XY 190 FC decreases to its starting weight. In comparison to its starting weight the sample weight of rhenus XT 46 FC is increased by 0.06 %. But in respect to the test’s standard deviation the weights can be evaluated as equal. A wash-down with demineralised water can be evaluated as being sufficient to clean work-pieces contaminated with PF.

In the ultrasonic bath the sample’s weights decrease more. It can be assumed that low masses of dust adhesions resulting from the milling process are removed by ultrasonic treatment. The samples treated by PF lose weight until 99.8 %. The weight of the reference samples even decreases to 99.7 %. The comparatively small weight loss of the PF treated samples verify the good flushing action of the PF: The mass of the dust adhesions is lower after wet machining than after dry machining. A consecutive butanone bath influences the sample weight within the test’s standard deviation. Therefore the influence of a butanone bath can be neglected.

In summary the sample weights only change within a range of 0.3 % by storing them in PF, cleaning them in demineralised water, ultrasonic bath and butanone bath. A wash-down with demineralised water is sufficient for rhenus XT 46 FC and rhenus XY 190 FC. In an ultrasonic bath very small dust adhesions are removed. Furthermore the special surface cleaning effect of the PF could be proved.

Summary

The objective of the conducted transverse tensile tests was to prove an effect of processing fluids (PF) on the cohesion of carbon fibres and epoxy resin. For this purpose CFRP samples were milled by applying PF and were stored for 168 h at 60 °C in the PF. Transverse tensile tests were conducted at 23 °C and - 50 °C to evaluate possible temperature effects at typical aviation conditions. As reference dry machined CFRP samples were used.

The testing results show that the effect of the tested PF rhenus XT 46 FC and rhenus XY 190 FC is very small and nearly not detectable. Except of one value the determined material properties are comparable with those of the dryly treated reference samples. According to the DLR, the effect of these PF in practice will be much lower in consideration of the tested worst-case-scenario. The real machining time and PF’s temperature are substantially lower.

From the examination of various surface cleaning processes resulted that a wash-down with demineralised water is sufficient for removing residues of the water miscible PF rhenus XT 46 FC and the water soluble PF rhenus XY 190 FC. Thus the tested PF are proved to be suitable for machining composites. |

Notes

Thanks

Special thanks go to the Institute of Compo site Structures and Adaptive Systems of the DLR in Brunswick for preparing, conditioning, and testing the test specimens as well as to the Institute of production technology of the University of Applied Sciences Zwickau for conducting machining tests.

References

  1. [1]
    Biermann, D.; Kirschner, M.: Wahl des Kühlschmierstoffkonzeptes als Schlüssel zum Erfolg. In: Werkstoffe in der Fertigung 5 (2015), S. 26 ffGoogle Scholar
  2. [2]
    Forbes, A.; Wittig, M.: Leichtbaubearbeitung im aktuellen Überblick. Online: www.maschinenmarkt.vogel.de/compositesworld/articles/436017/index4.html, access: 11 May 2016
  3. [3]
    Berufsgenossenschaft Holz und Metall, Bearbeitung von CFK Materialien — Orientierungshilfe für Schutzmaßnahmen, Ausgabe 10/2014Google Scholar
  4. [4]
    Beckenlechner, R.; Herbst, BH.: Kühlschmierstoffe für die spanende Bearbeitung. In: Lightweight Design 5 (2015), S. 38 ffCrossRefGoogle Scholar
  5. [5]
    Demmerling, A. L.; Schlindwein, H.-J.: Processing fluids for machining of carbon fibre reinforced plastics. 7th International Symposium on Composites Manufacturing for High Performance Applications (Braunschweig), 25.08.2016Google Scholar
  6. [6]
    Hofacker, A.: Composites mit Spezialkühlschmier stoffen effizienter bearbeiten. Online: www.springerprofessional.de/verbundwerkstoffe/leichtbau/composites-mit-spezialkuehlschmierstoffen-effizienter-bearbeiten/12165364, access: 28 April 2017
  7. [7]
    Kleineberg, M.; Nickel, J.; Pabsch, A.; Sigle, C.; Schöppinger, C.: Equipment and method for production of fibre-reinforced composites and/or plastic components according to a modified RTM method. Patent DE 198 53 709 C1, 2000Google Scholar
  8. [8]
    Erhard, G.: Konstruieren mit Kunststoffen. München/ Wien: Carl Hanser Verlag, 1993Google Scholar
  9. [9]
    Otto, M.: Chemometrie, Statistik und Computer einsatz in der Analytik. Weinheim: VCH, 1997Google Scholar

Copyright information

© Springer Fachmedien Wiesbaden 2017

Authors and Affiliations

  • Anna Lena Demmerling
    • 1
  • Hans Jürgen Schlindwein
    • 1
  • Sebastian Geier
    • 2
  • Thorsten Mahrholz
    • 2
  1. 1.Rhenus Lub GmbH & Co KGMönchengladbachGermany
  2. 2.German Aerospace CentreBraunschweigGermany

Personalised recommendations