Skip to main content

Mechanical Modelling of Powder Compaction

Due to Corona is not possible to attend this defense in person, instead attend via this link: Pin code: 2139

To connect to the virtual room, use one of the following methods:

Method
Address
Web browser
https://vconf.kth.se then type vRoom_9312_record
SIP
vRoom_9312_record@vconf.kth.se
IP
130.237.210.139 and choose room 9312
Telephone
+ 46 8 790 6421 and choose room 9312

Time: Fri 2020-03-20 10.00

Location: Live streaming, (Swedish)

Subject area: Solid Mechanics

Doctoral student: Hjalmar Staf , Hållfasthetslära

Opponent: Professor Göran Frenning, Uppsala universitet

Supervisor: Professor Per-Lennart Larsson, Hållfasthetslära

Export to calendar

Abstract

Cutting tool inserts, for instance used in steel machining, have the requirement to be toughand are therefore most often manufactured out of cemented carbides, using powdermetallurgy. Manufacturing components with powder metallurgy has its advantages in highproductivity and good net shape. The powder is spray dried and compacted to half itssintered volume. Because of friction between the powder and the pressing tool, the densityafter compaction is uneven, leading to uneven shrinkage during sintering. To get the rightshape after pressing and sintering, the pressing tool must often be compensated, which isboth expensive and time consuming. By doing computer simulations of the manufacturingprocess, the shape after sintering can be predicted and used to compensate the pressing toolbefore it is manufactured, thus saving both time and money. Also cracks and porosity in thepowder blank can be predicted with such simulations.

This thesis studies mechanical modelling of powder compaction in general and compactionof cemented tungsten carbide powders in particular. Because of the amount of powdergranules in a typical geometry, the mechanical behavior is modelled with a continuumapproach, using the finite element method (FEM). Accuracy is important in the presentapplication and therefore a detailed elastic-plastic material model with a density dependentyield surface of Drucker-Prager CAP kind is used.

For accurate material modelling it is important to include relevant features and to excludeunimportant physical effects. In Paper A sensitivity studies are therefore performed inorder to conclude which properties in the material model that have a significant influence onthe result. The studies show that anisotropy can be disregarded in the current application.

In Paper B the effects from creep and compaction speed are studied. It is concluded thatcreep has no influence on the density after compaction, which also is confirmed by densitymeasurements using a neutron source in Paper D. The compaction speed on the other handinfluences the friction coefficient between powder and pressing tool, lower at increasedspeed. In Paper C frictional behavior is scrutinized experimentally with the aid of aninstrumented die. The friction coefficient is determined and analyzed, and it is shown that itdepends on the normal pressure.

The sensitivity studies in Paper A show that measurements of the local density are neededin order to determine and verify material properties. Since the analyzed powder containstungsten (W), which has a high atomic number, a polychromatic beam of thermal neutronsis needed. In Paper D it is shown that the local density can be measured with 3D imagingand a thermal neutron source.

From the results and conclusions in the above-mentioned papers, a material description forpowder compaction is suggested. This description is implemented in FEM in Paper E andapplied to reverse engineering in order to determine important material parameters.Experiments in a pressing machine with a pressing method that includes multiple unloadingsteps is used. The material description with the determined parameters is verified withdensity measurements using a neutron source.

urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-268185