chem reactions engine

[Daniel Damelin]

Hi folks,

I did some testing of the current state of interactives that rely on the new changes in the chemical reactions engine. In general things look good.

One thing that looks like odd behavior when there is no heat bath is the repeated forming and breaking of the bond. Run this model and stop it just after a bond forms and then step the model back and forward to see the bond form-break-form …:

http://lab.concord.org/branch/chemical-reactions-energy-conservation/interactives.html#interactives/sam/chemical-reactions/1-what-is-a-chemical-reaction.json

I think this behavior is OK for now, but there was some discussion about how to avoid this which we should explore soon. I don’t think it should hold up the release or merging of the code into master.

There are a couple of ITSI interactives, which, regardless of old or new code, seem to be missing a heat bath.

http://lab.concord.org/branch/chemical-reactions-energy-conservation/interactives.html#interactives/itsi/chemical-reactions/dissocation-energy.json

http://lab.concord.org/branch/chemical-reactions-energy-conservation/interactives.html#interactives/itsi/chemical-reactions/dissocation-rites.json

If these were intended to not have the temperature of the model held constant then they need something like a temperature graph and buttons that increase and decrease the temp. As it stands now the temperature slider implies that it is showing the current temperature, but instead it just sets an instantaneous temperature which will change change once the atoms form or break bonds, unlinking the actual state of the temperature slider from the state of the model temperature.

However, this is not really related to current release, just something the ITSI folks may want to keep in mind.

-Dan