all the chassis I have done have been 2 part and I certainly wouldn't do a 3 part chassis (every join is a weakness) but it's up too you.
several ways too join but by far the most popular is tounge an groove glue join
Normally the bottom mould has the groove and the top mould the tounge
as a generalization let's say we have a tub with a finnished thickness off 14mm
2mm thick skins either side off 10mm core.
this is also on the premise that most tubs are oven or clave 3 stage cure.
so on the bottom mould before you actually make a part you need too make a 10mm thick "spacer"
( and cure it) round the top off the mould,this needs too be offset 2cm from the mould surface and deeper than the planned depth off your groove so it can be cut too you actual groove depth.
when you make the bottom off the tub you laminate the outer skins too the top off the mould an cure.
you then do your core but rather than going all the way too the edge off mould you put your pre made spacer in too replace the last 30mm or so of core and cure.
when you do the inner skins you lay up over the core and the spacer too lever with the top off the mould then cure
after cure you remove the spacer and you have your "groove"
for the "tounge" before you cure you lay up a 2cm skin on the top off the mould deeper than your "groove" and cure
after cure needs trimming too the same depth as your "groove"
this loose tool is fitted too the mould before you laminate and you lay you skins up over it too the top off the mould an cure
when you do your core,where your loose tool is you use 6cm core not 10 as the 6cm with 2cm skins either side will give you your 10mm "tounge".
once the core is cured you can do your inner skins
two halfs should fit together fairly well with a bit off tweaking,you fit your bulkheads as you go and bond up all in one go

You may find it more realistic to include a semi-structural space frame in the front section of the car.  You will need a space frame for roll bar backwards anyway (most road engines arent designed for structural use like in F1), so it might be easier to extend this forward around the cockpit and to include suspension pick ups up front.  If you then do all the roll bars, seat and harness pick up points to MSUK safety rules as in the Blue Book, then you can reassure anyone who might ask on the safety front.  Also means you will have the basic underpinnings to make it safe for use in some events like hillclimbs and some sprint series.  

Although as they have recently introduced engine/power limits to novice licence holders, so you may have to drive something lame to get enough signatures to go unlimited.

Warren Penalver
Easy Composites / Carbon Mods - Technical Support Assistant

Honeycombs are often bonded together with an adhesive film under vacuum and heat. Generally such a tub would be almost all pre-preg and done in multiple shots to get the honeycomb core.   

In theory you could do infused inner and outer skins then bond it all together. Your parts would all need to be postcured to a temperature they can survive the film curing cycle.

Alternatively you would have to manually apply adhesive and build it in stages -  perhaps with a vacuum bag to hold things in place.

It would be far easier to do an infused foam cored layup if you can make the design strong enough.

I imagine it would be a nightmare to do and would be a lot of effort.  YOu would end up bonding the tub together so you would need properly designed joggle joints and overlaps in the correct places.  Then you need to consider hard point inserts, additional bulkheads etc etc etc all as part of the original design.  Often the assembly would need to be done in a set order to make sure you get all the items in place at the right place and time in the process.

The other factors to consider are safety and race regulation.  If you are racing this, often there are a lot of series and MSUK regulations over composite structures and safety.  You are likely to need rear and possibly front roll hoops in the specified steel grades and design.  Plus often they specify limits on construction materials and/or additional layers such as an aramid/para-aramid type fibre for impact and anti-penetration purposes.  Plus of course a front and probably rear crash structure which will have to meet its own set standards for energy absorption.

If its track day only, you still need to make it safe for you and others around you but that is more or less down to your own concience in most countries as there is often less or no regulation on "construction" of cars for track days as long as it passes very basic scruntineering.    Although be aware many track day operators do not allow open wheeled cars except on specific open wheel car days or on "test days".  



Oh,  there's a few current and former F1 composites experts on this forum who can probably go into hours worth of extra detail on this subject should it be needed. 



Warren Penalver
Easy Composites / Carbon Mods - Technical Support Assistant

Good day,  I am in the planning stages of constructing a monocoque for a F1 dublicate utilizing a 1000cc motorcycle engine (gsxr/zx,cbr...) it will be producing about 3-500hp being built and turbocharged.  I am working out the construction stages of building the monocoque and I will first be building a plug using pink panther insulating foam afterwards will be a 3 part split mold covering from the front to the back all the way up the middle then the upper piece will have two splits along the center line. I plan on using 8 layers of carbon with nomex 1/2" core sandwiched between them also it will have kevlar sandwiched between the two layers of carbon on either side of the core ex: C/C/K/C/C/Nomex/C/C/K/C/C.  I was planing to vacuum infuse the parts separately and will have to be done in stages with the first cloth layers being done then bonding the nomex core with wet layup after that finishing it with the rest as you can't infuse resin with a honeycomb cores.  My problem is how would the 3 section be finally bonded together and strengthen as it would need an overlap or additional layers where the bond would be but being a monocoque it would be difficult to reinforce it properly by the frontal area.