Automatic result testing on Travis

[mrp089]

Run automatic test cases and compare them to analytical or numerical solutions (on Travis). Create a minimal test case for each feature of the solver and compare it to a previously computed solution. Whenever possible, use analytical solutions as a comparison.

If features are not tested or found to be defunct, remove them from the Code, input files, and documentation entirely.

Please modify the following lists accordingly:

• add missing features
• strikethrough removed features
• check tested features

Testing framework

• script to automatically run test cases with builds of svOneDSolver
• script to compare simulation results to stored results up to a given tolerance

General features

• NODE
• JOINT

Loss types

• NONE
• STENOSIS
• BRANCH_THROUGH_DIVIDING
• BRANCH_SIDE_DIVIDING
• BRANCH_THROUGH_CONVERGING
• BRANCH_SIDE_CONVERGING
• BIFURCATION_BRANCH

Outflow conditions

• NOBOUND
• PRESSURE
• RESISTANCE
• RCR
• CORONARY
• AREA
• FLOW
• RESISTANCE_TIME
• PRESSURE_WAVE
• WAVE
• IMPEDANCE
• LPN
• ADMITTANCE
• PULMONARY
• CLOSED LOOP

Inflow conditions

• FLOW
• PRESSURE_WAVE
• AREA
• RESISTANCE
• RESISTANCE_TIME
• WAVE
• RCR
• CORONARY
• IMPEDANCE
• PULMONARY

Formulation Type

• Advective
• Conservative

Stabilization

• No stabilization
• With stabilization

Material Type

• LINEAR
• OLUFSEN

Output

• TEXT
• VTK

[caseyfleeter]

Updated the minor losses issue (#23) to reflect the need for Travis testing. Removed those (strikethrough) from this list in the meantime, as they are not currently included in the code.

[caseyfleeter]

Updated the boundary condition issue (#22) to reflect the need for Travis testing. Removed those (strikethrough) from this list in the meantime, as they are not currently included in the code.

[melodydong]

The advective formulation seems to be fairly sensitive to the boundary conditions and timesteps. We might want to only support the conservative formulation.