In this work, the temperature distribution on an industrial mold tool is monitored during autoclave runs with three settings. In one of the settings, the temperature and pressure follow a scheme used in real moldings, while in the other two cases, the temperature is increased as fast as possible with and without an applied pressure. The temperature difference over the tool is relatively large and varies between 29℃ and 76℃ validating a detailed investigation of the temperature at different points. Two results of this are that positions on the up-stream side of the tool are heated faster than positions down-stream and the heating over the tool is symmetric while that within is asymmetric. Roughly estimated heat transfer coefficients reveal that the temperature ramping has no significant effect on the local heat transfer coefficients while the applied pressure more than doubled them. In addition flow field measurements with particle image velocimetry are performed, revealing a very slow flow near the roof of the autoclave and a velocity peak near the floor of it, indicating that the flow profile within the autoclave and variation in heat transfer coefficients should be considered in autoclave simulations.

In this work, computational fluid dynamics simulations are performed to predict the temperature distribution on a part during an autoclave run. Data from an experimental study are used as input to the simulations and also for comparison with the numerical results. A conjugate heat transfer approach was used for the simulations, where best agreement with experiments was obtained from the simulation that included thermal radiation and utilized an experimentally obtained velocity profile as inlet velocity. A yet more detailed inlet velocity profile and more advanced turbulent model could result in an even better agreement.

Compression moulding experiments of sheet moulding compound, visual observations of a vacuum test with prepregs and numerical models with two main approaches for computational fluid dynamics simulations of the mould filling phase are presented. One assumes that there are layers near the mould surfaces with much less viscosity and the other only use one viscosity model. The numerical experiments showed that the pressure could be accurately predicted with both approaches. The property necessary to predict correct pressure with altered mould closing velocities was that the bulk material had to obey shear-thinning effects. Preheating effects before compression were neglected, but altering the heating time until the prepreg was assumed to start flow had a significant effect. The experiments confirmed that the pressure is predominantly affected by the mould closing velocity. Regardless of the considered process settings, a first pressure top always appeared approximately at the logarithmic strain 0.25. A second top was associated with a slowdown of the press. The location of this was affected by the velocity and the vacuum, the latter indicating that vacuum assistance prevents a build-up of back pressure. Furthermore, heated prepreg above a critical temperature is observed to swell immediately as vacuum assistance is applied.

The work presented in this thesis focus on the flow taking place in two industrialmanufacturing methods for composite manufacturing. The first method underinvestigation is compression moulding of Sheet Moulding Compound (SMC).With this method, SMC prepreg are stacked and placed in a heated mould toolwhich then is closed and held under pressure until the part has cured. Thismethod can be used to manufacture large quantities since the cycle time is short.The drawbacks with the method include pores at the surface causing largervisual defects such as blowouts if the part is painted and relatively poormechanical properties as compared to composites made out of continuous fibres.The aim of this work has been to increase the understanding of the flow that isgenerated during the filling of the mould and how this affects defects such aspores. This has been done experimentally with a circular moulding toolequipped with pressure sensors and vacuum assistance capability. The utilisedpress allowed different temperatures and closure velocities to be used. Thequality of the moulded experimental SMC plates was quantified with differentmethods. Pressure sensors revealed the pressure inside the mould and the flowbehaviour could for example be analysed with image analysis of plates mouldedwith multi-coloured SMC. The relative void content was measured with a highvoltage insulation test. The experiments showed that SMC that had flowed alarger distance had less voids than SMC that still was in the centre.Investigations of the coloured SMC indicate that a high closing velocity (10 mms-1) of the mould gives a more homogenous flow and also that by applyingvacuum assistance, a more homogenous flow is achieved with low closingvelocity (2.5 mm s-1). Interestingly the settings that resulted in a morehomogenous flow also resulted in less voids, indicating that a more homogenous flow should be sought for if a low amount of voids is desired. The setting that gave overall best result was to use vacuum assistance (75% vacuum), the low mould temperature (144 °C vs. 154 °C) and low velocity. Vacuum assistance also seemed to prevent a back pressure inside the closed mould tool since the press was able to compress further with constant velocity with vacuum assistance. In addition to the experimental work, non-Newtonian viscosity models for SMC have been developed for Computational Fluid Dynamics simulations which could predict the pressure at different closing velocities. The models are complex, but the one thing that seemed to be most important in order to predict accurate pressure at different closing velocities was to allow the bulk material to behave shear thinning. Along with the models, a method of finding unknown constants is presented which allows the models to be used for various SMCs without knowledge about their real material properties. In addition to the experimental work with SMC, the behaviour of bubbles in a non-Newtonian fluid during compression was investigated experimentally with Particle Image Velocimetry and the bubble motion was furthermore also analytically modelledand the results were coupled to the experiment results.

This work, that involves both experiments and numerical simulations, concerns autoclave molding. An autoclave is basically a pressure vessel, where the entrapped and often highly compressed gas is heated and circulated in order to heat the components that have been placed inside the vessel. In the autoclaveprocess, the desirable state would be that an even and optimal temperature existed in the whole part that is manufactured. Unfortunately, this is not always the case. All in all we need to get a better understanding of the flow inside an autoclave and the convective heat transfer from the heated gas to the composite components. In this work we have therefore investigated the flow behavior by performing qualitative measurements with particle image velocimetry inside an autoclave. The concept is to dope the gas within the autoclave with smoke and illuminate the smoke with a thin sheet of laser light. Captured images of the moving smoke are then cross correlated to give velocity fields. We have also investigated the heat transfer to the tool by measuring the temperature at multiple locations during heating. The obtained velocity field is used to produce inlet condition for the simulations, performed with Computational Fluid Dynamics, which subsequently are compared with the experimentally obtained tool temperature. The simulation technique may then be used to optimize both the tools, and the actual location of the tools inside the autoclave in order to improve quality and reduce costs.

På grund av utmärkta materialegenskaper och relativt låga material ochtillverkningskostnader har användningen av polymera kompositmaterial ökat under de senaste årtiondena. En tillverkningsmetod som lämpar sig för storskalig produktion, t.ex.lättviktiga fordonsdetaljer, är pressning av Sheet Moulding Compound (SMC). Trots att denna tillverkningsmetod har förbättrats avsevärt sedan den introducerades återstår fortfarande vissa problem som måste lösas. Den huvudsakliga anledningen till varför SMC inte kommit till större användning inom fordonsindustrin är otillfredsställande kvalitet på ytor på tillverkade detaljer på grund av porer. I det här arbetet har experiment och numeriska simuleringar genomförts för at öka förståelsen om hur materialet flödar under tillverkning och hur flödet påverkar kvalitén på färdiga detaljer med avseende på porer. Processparameterexperiment genomfördes med SMC för att undersöka effekterna av vakuumassistans, verktygstemperatur och presshastighet på flödesbeteende och portransport. Den relativa porhalten kvantifierades med hjälp av ett elhållfasthetstest och det slutgiltiga flödesmönstret studerades med hjälp av färgat SMC. Sammanfattningsvis visade de experimentella resultaten att inställningen med vakuumassistans, låg presshastighet och låg verktygstemperatur skapar ett homogent flöde och minimaliserar porhalten.För att ytterligare öka förståelsen av portransport genomfördes även ett modellexperiment där en icke-Newtonsk fluid (fett) med tillsatta luftbubblor pressades mellan två transparanta plattor och studerades med hjälp av Particle Image Velocimetry. Experimenten visade att hastigheten för luftbubblorna ökade mer än hastigheten för det omgivande fettet under pressning. Under det senare stadiet av pressningen när luftbubblorna accelererade ändrade de form ifrån att i början ha varit ungefär sfäriska till den karakteristiska formen av en fallande regndroppe. Luftbubbelrörelsen modellerades även analytiskt och den utvecklade analytiska modellen stödjer den visade utvecklingen i experimenten.Slutligen användes den kommersiella mjukvaran Ansys CFX till att genomföra modellering av flödet av SMC under tillverkning. Två olika viskositetsmodeller testades där det modellerade trycket i två punkter jämfördes med experimentellt uppmätta värden. I den första modellen sågs SMC-flödet som ett rent töjningsflöde där viskositeten berodde på temperatur, fibervolymhalt och töjningshastighet. Medans flödet i den andra modellen även sågs som skjuvförtunnande närmast verktygskanterna men som ett töjningsflöde i övrigt. Av de två modellerna verkar det som att den senare modellerar trycket bäst jämfört med experimenten.

Bubble transport is of importance in many applications for non-Newtonian fluids, such as in the manufacture of composite materials where residual bubbles may impair electrical properties, surface appearance and the mechanical properties of the finished products. In this study, a model experiment is performed where a non-Newtonian fluid (grease) containing bubbles is compressed between two plates, whereby the motion of the bubbles is tracked and evaluated using Particle Image Velocimetry. The bubble motion is furthermore analytically modelled and coupled to the experimental results.

Bubble transport is of importance in many applications for non-Newtonian fluids, such as in the manufacture of composite materials where residual bubbles may impair electrical properties, surface appearance and the mechanical properties of the finished products. In this study, a model experiment is performed where a non-Newtonian fluid (grease) containing bubbles is compressed between two plates, whereby the motion of the bubbles is tracked and evaluated using Particle Image Velocimetry. The bubble motion is furthermore analytically modelled and coupled to the experiment results.

Due to excellent properties and relatively low material and manufacturing costs, the use of fibre reinforced polymer composites have increased during the last decades. One method that is suitable for large scale productions of e.g. lightweight vehicle components is compression moulding of sheet moulding compound (SMC). Although the technique has been considerably improved since it first was introduced, some further improvements need to be done. The main reason why it has not come in wider use in the vehicle industry is unsatisfactory conditions of the surface finish of parts manufactured due to voids. In this work, experiments and numerical simulations has been performed in order to increase the knowledge of the flow behaviour during the compression moulding process and how the flow affect the quality of the finished product. A process parameter experiment of the compression moulding phase, carried out with a design of experiment approach, was performed in order to investigate the effect of vacuum assistance, mould temperature and ram velocity on the void transport and flow behaviour for SMC. The relative amount of voids has been quantified with a high voltage insulation test and the flow behaviour has been quantified with image analysis of samples moulded with coloured SMC. In conclusion, the setting of high vacuum, low ram velocity and low mould temperature creates a homogeneous flow and minimises the amount of voids. In order to further increase the understanding of void removal during compression moulding, a model experiment was performed where a non-Newtonian fluid (grease) with added bubbles was compressed between two plates whereas the motion of the bubbles were tracked and evaluated using Particle Image Velocimetry. The bubble motion was furthermore analytically modelled and coupled to the experimental results. The experiments reveal an increase in bubble speed compared to the surrounding grease during the compression of the plates. During the latter stage of the compression, the particles change form from initially being approximately spherical, to have the characteristic form of a falling raindrop. The change in form coincides with the increase in speed of the bubble. The developed analytical model supports the shown development in the experiments. A full general solution comprising an arbitrary value of the Power Law exponent, for the velocity fraction coefficient representing the relative bubble speed, is however not covered at the present stage. Finally, the commercial software Ansys CFX were used to perform computational fluid dynamics (CFD) modelling of the flow during compression moulding with a two different multiphase models. The first model treats the flow of SMC as purely extensional and dependent on temperature, fibre volume fraction and strain rate. While the other one sees the flow as mainly extensional but also with thin shear layers near the surfaces of the moulding tool. Where the viscosity, in addition to temperature, fibre volume fraction and strain rate, also is dependant on shear strain rate. Of the two models, the latter seems to be more robust in modelling the pressure during moulding.

During compression moulding of sheet moulding compound (SMC), voids are formed that can deteriorate the properties of the final product. Here, experimental work and CFD-simulations have been carried out in order to increase the knowledge of the SMC compression moulding behaviour which highly affects the quality of the final products.

The effect of vacuum assistance, mould temperature and ram velocity on the void transport and flow behaviour for sheet moulding compound (SMC) have been investigated with a design of experiment approach of the compression moulding phase. The relative amount of voids has been quantified with a high voltage insulation test and the flow behaviour has been quantified with image analysis of samples moulded with coloured SMC. In conclusion, the setting of high vacuum, low ram velocity and low mould temperature creates a homogeneous flow and minimises the amount of voids.