Diorama resin

Most two-part epoxy resins have an exothermic reaction.   That is they generate heat.   The larger the volume the more heat   

Embedding a plastic model in a brick of clear resin is a sure fire recipe for melting the model, or at least severely warping it.  

Try mixing and pouring your resin in thinner lifts.   One quarter inch is about the best you can do

Alternatively,  build a box of clear plexiglass and suspend your sub inside.  Perhaps a thin layer of blue-tinted resin on the top as the water surface. 

Would it be possible to create a mould in clay or something, and then create a positive out of something that will withstand the heat of the resin, then replace the positive with the real thing? 

It's not clear if you were showing a sub surface running on a sea, or trying to show it submerged.

If the latter, suspending it in a block of resin will not work.

In no particular order:

Very costly.

Heat as you mentioned.

Pouring in lifts it's impossible to eliminate bubbles between layers, which spoils the effect.

Building a shadow, or perspective controlled, box that's air filled under a sea surface works.

Also be aware that in WW2 a subs maximum depth wouldn't be much deeper than it's length. Since then not much better.

What Morrison said,  but lets examine your premise.   

As a though experiment this is how I'd go about it.   I'm supposing that you want to imbed  a 1:700 scale submarine in a brick of clear epoxy resin.  Larger sizes will drive more cost or differnt material choices.  That size presumes that the two brick half sizes will be 3 to 6 inches deep.    I'm going to provide info for Smooth-on resins.    I have used their RTV and resins and am familiar with them

Their Crystal Clear product lists item 204 as the 3 to 6 inch deep material.   See the videos.   Suppose that you are going to use a method similar to the one shown wherr they pour a clear sign with back cut lettering.  

https://www.smooth-on.com/product-line/crystal-clear/

Here is the technical info on the material.   Note the caution that it is for industrial use and may require a respirator for use.  Toxic if inhaled.

https://www.smooth-on.com/products/crystal-clear-204/

And here is the material safety data sheet

http://www.smooth-on.com/msds/files/480A_1_481B_1.pdf

I followed the purchase links and IIRC it runs about 200 dollars a gallon!  Other required materials include modeling clay, RTV, mold release, and mold box materials

Make your negative mold from clay.   This is in a mold box the the size of the finished project.  Lay  modeling clay 1/2 the depth of the box. Imbed the submarine model halfway.   The surface of the clay MUST BE perfectly smooth.  No ridges or fingerprints.  Pour RTV rubber.  Allow to harden.   Disassemble the box.   Pull the all the clay, clean the model and re-imbed it in the RTV.  Reassemble the box.  Coat the RTV & model with mold release.  Mix & pour more RTV.   Allow to harden.   Diassemble the box.   Separate the two RTV sheets, front & back. Pull the model & set aside.

Reassemble box.   Put one of the RTV sheets in the box.   Liberally coat with mold release.   Mix & pour the clear resin.   Recommended process may include pressure/vacuum ro elinimate bubbles.    Allow to harden.   Note that Clear Cast 204 is 48 hours.    Be patient.   Rush it now and you WILL f**k it up.

Diassemble box.   Pull the half brick.   Replace the RTV sheet with the opposite.  Reassemble the box.   Apply mold release.  Mix and pour clear resin. Pressure/vacuum as needed.   Allow to harden.

Disassemble box.  Pull half brick.   Place model in reserved space between 2 half bricks.  Mix a bit of the clear resin to act as glue.   Brush on one of the brick surfaces and assemble the bricks, hope for some but not a lot of squeeze out.  Clamp and allow to harden.   

Polish the join surfaces.

That is my thought experiment on how I'd approach your desired method.    Smooth-On also has a pretty good technical assistance desk.   Perhaps check with them for their opinion

Just a quick thought regarding the use of Clay with both RTV and resin - read the labels carefully. Some clays contain traces of sulfur, which doesn't necessarily play well with some resins and molding rubbers. In some cases, the RTV won't set up properly and the mold turns to a gooey mess where the latex/silicone/rubber has leached up the sulfur. In other cases, the resin gets affected and you wind up with soft, rubbery bits that never firm up properly.

Read the manufacturers' instructions and safety data sheets when you get to the point where you are ready to begin shopping for materials. They're normally full of information that you'll find useful, including a warning if the material is reactive to sulfur and the amount of heat you can reasonably expect during the reaction. 

Also, if making any sort of mold of the sub kit, remember that anyting resbling rails, periscope shears, seats on guns and the like is like to be mashed or bent using clay.  Or ripped clean off if using something like RTV.

I have heard that using several clear coats, especially acrylic clear coats can help protect the plastic.  But, you really need to use shallow "lifts" of the casting resin with a lot of drying time in between.  And cost you an arm and leg for the cost of a the quarts of resin required.

Really, clear plexiglass (or acrylic sheet) box works far better.  You can paint the back to similate depth, and the top, can be modeled as the surface.  And, you can better model things like WWII boats seldom going deeper than 100m/300', and usually not going more than half that-- a very managable 65mm/2.5" down at 1/700, and near any amount below that, as the oceans average 3600m/12,000 deep.

Paulv718

I want to put a submarine in resin in a box to make it look like water. I used an epoxy and my model melted. Any suggestions on a good resin for this. Thanks in advance 

I second Smooth-On, their whole product line is excellent, and easy to use.

Also, you might find this video useful.  The modeler presents your subject, a submarine under water:

https://www.youtube.com/watch?v=PM_-WBmbNJc