i would go for steel rings shovel and knife!!! are you then using pressure plates for water retention curve
Fitting van genuchten and then Mualem using Hydrus or equivalent? check out my questions in relation to pressure plates and useful answer by many.
Also consider texture - sand silt and clay, bulk density and particle density. These will help you define better your fit and also give you total porosity.
In loess area located in Gansu Province, China, we use a 100 cm3 steel rings to take three undisturbed samples for field water content and density. Then, these samples were used to determinate special gravity (this can calculate total porosity, along with dry density), and grain size analysis. So, this can gain many basic properties of soil, as suggested by Owen Fenton.
For saturated hydraulic conductivity test, a bulk sample or 3-4 steel ring sample (penetration test standard) were taken. Our laboratory (Lanzhou University) is very close to the study area, so we take sample using steel ring. You could consider a bulk sample method. But you must deal with the sample in-situ carefully, for eample, wax-sealed.
You could consider the suggestion from Dr. Owen Fenton, use pressure plate to test SWCC of soil. Because many question associated with soil erosion may be related to unsaturated soil.
Additionally, physico-chemical behavior and aggregate feature are very important to soil erosion.
I would be more convinced of the accuracy of your bulk densities (and the minimum disturbance in your cores) if you use a cylinder of 100 mm diameter and 100 mm length, i.e. about 785 cm3, rather than the 100 cm3 cylinder. The 100 cm3 cylinder is about 50 mm in diameter and 50 mm in length. In my experience it is difficult to maintain the correct orientation of the 100 cm3 cylinder while knocking it into the soil (although I concede it may be easier in some soils such as loess-derived soil). Furthermore, with bulk density we are interested in small differences such as the difference between 1.60 and 1.61 Mg/m3, i.e. a difference of 0.7%. It is critical, therefore, to get the volume accurate in the calculation of mass/volume. Mass of the contents of the cylinder is easy to obtain accurately. Volume is the problem. Even if you measure with a vernier caliper the length of the cylinder at 8 points and the diameter of each end of the cylinder at 4 points, there is still a lot of uncertainty in the volume. This together with the cylinder not going in quite straight plus some edge effect where the soil meets the cylinder wall makes for a rather unsatisfactory bulk density. The uncertainty is greatly reduced using 100 mm dimensions.
At University of Ibadan we used bulk density cylinders of 100 mm dimensions from a soil research supplier. But I believe you need to ask around water companies etc to find some wide-diameter steel pipe as we did successfully here in Mozambique.
Concerning hydraulic conductivity, please check the papers by Gumbs and others which showed that field-measured soil hydraulic conductivity differs e.g. by a factor of 10 from values obtained using cores. So field-measured values may be more meaningful.
The problem with cores is shattering during hammering and compaction along sidewalks. The only way to really reduce this is bigger cores I use either 65 mm or 100 mm diameter cores, with similar width to length. Don't use long cores ie 65 mm x 120 mm they have maximum sidewall error.
If you want to measure infiltration, soil water properties do as much as you can in the field and insitu. Look at my paper on determining k sat, values differed up to FOUR orders of magnitude depending on methodology. Paper can be downloaded from my researchgate site.