Recognizing that the posting party feels that FDM constructed parts are of insufficient strength for his purpose and allowing for proper layer bonding, one can understand that the model can be perfectly constructed and not reach the strength objective.
Filling a model with an epoxy or a casting resin will provide additional strength. Testing smaller, non-critical models is recommended to determine the level of increase. The design has to be re-engineered to provide for resin/epoxy flow within the model. Some epoxy and resin formulations generate heat when curing and may soften the model. The solution in such cases is to mix and pour small amounts, allowing for a pause between pours.
An alternative to filling a model is to reduce the perimeter (if applicable) and apply a reinforcing layer. I've constructed satisfactorily printed models with insufficient strength for my purposes, but then applied fiberglass cloth and epoxy resin to the outside. A single layer provided the necessary strength in my case.
One could apply sufficient layers to provide all the necessary strength, effectively turning the model into a positive mold. This is commonly done with amateur-built aircraft of the Burt Rutan design style. More recently, an article appeared on the internet of a model being printed with wash-away support material only, covered in carbon fiber. The wash-away was washed away and the wing structure became the product. For your application, it may not be necessary or practical to remove the inner model.
Just as with the injection concept, one must re-engineer the model to allow for this type of reinforcement. Edges will have to be radiused or the sharp termination of reinforcement layer will become a weak point. Tight inside angles will have a similar problem.
Fiberglass cloth comes in various weights, measured in ounces per square foot (US). The lighter cloth is more capable of "turning corners" and fitting into tight angles.