Hello Ted:
You are on the right track with your proposed steel bedplate & slab. As Limy Sam correctly notes, the portion of the bedplate which extends to mount the outboard bearing has to be every bit as rigid as the rest of the bedplate. Personally, I would simply make an "L" shaped bedplate, not going with an "A" shaped extension. While the triangulation of the A shape will be quite stiff, I tend to think in terms of ease of fabrication. If you make your bedplate in an "L" shape with a "well" for the flywheel, I am sure it will be fine. 10 mm plate is approximately 3/8" ( I am a dinosaur who thinks in inches and does the mental math for rough conversions to metric and vice-versa). Depending on the size of the bedplate and the main structural members, and your noting the flywheel as weighing 2 tons, I am thinking I'd go up to at least 12 mm ( 1/2") plate and some hefty Wideflange beam sections (or whatever the UK equivalent of a US Wideflange is). Many years ago, engineers at Bethlehem Steel realized the traditional "I" beam sections were lacking structurally for many applications. They came up with the Wideflange series of beam sections. A traditional I beam section relies on the depth of the beam to space relatively narrow flanges far enough from the neutral axis of the I beam to get a high enough moment of inertia to handle structural loadings. I beams tend to be deeper and narrower and more 'twisty' than wideflange sections. I almost never use I beams in my structural designs for those reasons. Without knowing more about your Crossley engine, I am 'shooting from the hip, in the dark' in terms of what information or help I can provide you with,
I would also suggest that when you build your bedframe, you might plan on putting some heavy plate steel pads (say 25 mm or 1" plate) wherever the engine mainframe anchor bolts are located. These pads give you something to tap for the anchor studbolts, and give you something to shim off of. It has been my experience with engine mainframe castings that the underside of the casting was often left unmachined. This was based on the presumption that the engine was going onto a masonry (or concrete) foundation and grouting would be used to take up for irregularities in the bottom surfaces of the mainframe. I address this by using a soft copper shim on top of each shim pack at each anchor bolt location. The soft copper shim lets the rough as-cast surfaces on the bottom of an engine mainframe 'bite' into the copper. If Crossley was so thoughtful as to plane off the bottom surfaces of the mainframe of your engine, you are way ahead of the game.
As for welding, I would avoid putting continuous 'running' welds on the top of the fabricated steel baseplate. The intersecting weld joints where the main structural members (which I would be designing as wideflange sections here in the USA) get full welds. The flanges get full penetration welds, webs tie to webs at the intersecting connections with fillet welds. The top plate should be put on with a stitch weld to avoid excessive weld stress and subsequent distortion. I'd use a flux-cored/gas shielded welding procedure (known as GFCAW) as it puts down a lot of weld and has a lot less post-weld distortion than SMAW (shielded metal arc welding, aka "stick welding"). Stick is OK for small welds and tacking things in place, but to get the amount of weld put down you will be needing, I'd go with FCAW.
Since your engine will be sitting on a steel bedplate rather than a masonry foundation, it should be 'chocked'. On a traditional masonry foundation, engine mainframes were grouted. The grout mix took up any irregularities in the surface of the foundation as well as the underside of the engine mainframe. This grout also served to chock the mainframe from moving fore-and-aft (as would be the case with a horizontal engine) in relation to the foundation. On steel foundations, such as aboard ships, machinery was chocked against movement relative to the steel foundations it was bolted to. Chocking was done by welding or bolting steel blocks that bore solidly against the vertical edges of an engine's mainframe or other machinery's mounting feet. On smaller machinery, reamed and fitted dowels were used. In more recent times, chocking is done using epoxy resin compounds. These work great.... until you have to remove the machinery from its foundation. That is when you wish you never heard of epoxy resin choking compound and want to call in a tactical airstrike. In your case, welding some 'jacking dogs' to the top of the steel bedplate with jacking screws will suffice to chock your engine. The reciprocating mass in a horizontal engine wants to move the engine fore-and-aft and also can produce some lateral forces. Chocking with some jacking dogs and jacking screws (lock nutted) will take care of this matter and lets you remove the engine or tweak its final position on the baseplate.
Anchoring to the slab, I would suggest using drilled-in anchor bolts set in resin grout. Hilti and other firms offer a variety of resin grouts, some in capsule form, some in 'sausage form', and some in two-part cartridges for use in mixing 'guns'. No sense fighting with imbed anchor bolts. Get the slab in place and put the anchor bolts in afterwards. I'd also suggest chipping off the top of the slab's concrete before setting the steel bedplate. Chipping to expose aggregate will give a good bond for non-shrink grouting. I'd use a cementitious non-metallic non shrink grout and allow 1/2 to 1" of shim thickness between the underside of your baseplate and top of the slab concrete. The non shrink grout will take up any voids and give as close to complete bearing contact as possible. Once you have the bedplate grouted, you can fill it with what I call 'slush grout'. Non shrink grout is pricey and works well. For simply filling the bedplate to add mass dampening and cut out any 'empty drum' sounding from within the bedplate cells, we fill the bedplates with a sand-cement grout. This is something like 4 parts clean sharp sand to 3 parts of portland cement, mixed to a flowable consistency and poured into the cavities in the bedplate. We typically cut grout holes in the top plates of this sort of bedplate, allowing holes for trapped air to escape as well as to pour the grout into. The slush grout is brought right up to the level of the steel plate top on the bedplate and struck off flush with it. Makes a neat job and a couple of coats of enamel paint finish it so oil does not soak into the grouting.
Prior to placing the concrete for the slab, I'd suggest you put down a bed of crushed stone, perhaps about 2" on its largest side. If you are doing any grading or excavating for the area where the slab goes, a bed of crushed stone 6"-12" deep will let the slab rest on the soil and help with drainage under the slab since it sits on clay soil. I'd make the slab at least 10-12" thick to get some more mass and strength. I'd reinforce it with number 5 ( 5/8" nominal) rebars, two 'mats' tied on 12" x 12" centers, one mat at the top, one at the bottom of the slab. Allow 2" of cover concrete between the outer surfaces of the slab and the rebar. This cover concrete is important as it gets the rebar in deep enough in the concrete to get a good mechanical bond and do its work structurally. It also is better at resisting freeze thaw cycling. I'd spec the concrete with a 3-5% air entrainment, particularly if the slab is outdoors. The air entrainment helps the slab better stand up to freeze-thaw cycling. I'd use concrete with a minimum compressive strength (Fc') of 3000 psi, and pour with a stiff mix (a slump of about 3" max). Done a lot of concrete foundation design over my career and seen a lot of concrete placed for hydroelectric plant work, machinery foundations and building construction. Some up front work on the slab will insure your engine sits solidly and does not get up and dance when running, and will maintain good alignment between the main bearings and the outboard pedestal bearing.
