00:17:04 Kate Hendle: Welcome to Day 2 of the Cooling Environments Tech Talk! Please put any questions for the speakers here in the chat and keep yourself on mute if you are not speaking. 00:20:34 Archna Haylock: yesterday’s Cooling Env Tech Talk, along with all the other Tech Talks are available here: https://www.opencompute.org/events/past-events 00:25:37 Bruce Lv_Ragile Networks: Great! 00:26:42 Vishwas Patil: Is the Google drive accessible to all? 00:27:59 Reza Najjari (CoolIT): Last time I checked the Google Drive link, it was not allowing us to access the material database 00:32:01 Gaby Mindreci m-i2i.com: How about moisture content in the aging test? 00:41:22 Allison Boen: Will these fluids be designed by UL as “UL listed” In the document 00:53:59 Cosimo Pecchioli: in the "integrated solutions" are you considering the cooling part as well as part of the package? I mean, are you looking at a solution that includes hardware, software, tank, cooling and controls all in one package? 00:54:32 Cosimo Pecchioli: controls as a package? 00:58:32 John Nguyen: For the scale out for immersion topic, regarding "standards", is the goal to standardize monitoring and control of equipment (node and rack level)? 01:02:55 Mick Jones: Hi John, yes you should join the 11am Hardware Management for Liquid Cooling so we can capture any edge specific requirements. 01:03:25 John Nguyen: Thank you Mick, see you there 01:25:22 David G..: The core of embodied carbon of data center is the servers. Can liquid cooling reduce the quantity of the servers with the same IT task? 01:37:41 Gaby Mindreci m-i2i.com: What is the cost of one mated pair of the drip-free connector shown? You may state the cost in rough order of magnitude… 01:38:19 Gaby Mindreci m-i2i.com: Also, how many mating cycles is the connector designed to support? 01:40:07 Gaby Mindreci m-i2i.com: Also, how many mated pairs of these connectors are used per rack? 01:43:35 David G..: With 1” connection and 100LPM what is real rack load? 01:45:42 Gaby Mindreci m-i2i.com: If there is time for one more question: Are these connector capable to support blind-mating applications? 01:48:19 Gaby Mindreci m-i2i.com: But existing connectors today are about $80/mated pair... 01:48:24 John Fernandes (Infra): With 1” connection and 100LPM what is real rack load? It depends on your deltaT target and coolant of choice. For 10°C deltaT and PG25 as the coolant, that's roughly 67kW of liquid cooled load. 01:48:28 Paul Gwin (Intel): How does this connector differ from the UQD design work that was done previously. would that design scale to the needs you have …. low insertion force? Seems there has been a few attempts at UQD's - wondering how they compare? 01:48:42 Tejas Intel: Can this effort be included as part of UQD initiative? 01:49:15 Vishwas Patil: Sorry , I missed the earlier conversation over chat on the google drive access. Could anyone update me on that? 01:57:18 Noman Mithani: In regards to UQD: when this group was started UQD did not have any plans to support a large body quick connector since the group was focused on ITE and cold plate level connections. Since this large connector has different end goals and end users we decided to start a new group. 02:00:32 David G..: Is there only single phase cold plates? 02:01:20 Rob Curtis (Dell): Does the cold plate requirements cover interoperability between various vendors for cold plate loops, manifolds, CDUs, etc. For example allowances or requirements that would allow for mixed vendor solutions. E.g. cold plates from vendor A and vendor B, CDU from Vendor C, Manifold from vendor D? 02:02:22 Rob Curtis (Dell): Good to hear that. 02:19:15 David G..: What kind of CPU is used and what TDP per CPU is possible for that OCP server? Max. temperature of the liquid inlet/outlet? 02:19:57 Paul Gwin (Intel): What is the true position tolerance allowance of the valve connection alignment, wrt stress from misalignment? Does CTE and temp cycle create a fatigue stress that is concerning or is there temp cycle data on this? 02:19:58 Cosimo Pecchioli: Are you considering any heat reuse opportunity? 02:20:12 David G..: Great presentation! Thanks a lot! 02:21:22 Cosimo Pecchioli: perfect pronunciation of my name! :) 02:28:02 David G..: Beauty color. 02:32:07 Gaby Mindreci m-i2i.com: In some of these photos it appears that galvanic corrosion is the primary suspect. It may benefit if a material compatibility assessment is applied at the design phase and sacrificial anodes are installed in risk areas especially increase the life span of the structures. 02:32:32 Paul Gwin (Intel): In these corrosion pictures of the HEX, were there galvanic couples directly in electrical conduction contact (tough to see from pictures - looked like Cu in contact with Al)? 02:35:01 David G..: Do we need demand stainless steel frames for air cooled heat exchangers? 02:36:08 David G..: *air heat exchangers 02:44:44 Gaby Mindreci m-i2i.com: Interesting presentation, good day’s content! Thanks to all, have to drop… 02:46:08 Michael Gonzalez | CEJN : Have to drop off. Great Job Everyone! 02:46:58 Paul Gwin (Intel): WRT the joint porosity - do you suspect that is due to lack of process control? perhaps flux outgas during liquidous phase? 02:54:03 Troy: Need to drop off. Good work! 02:54:04 James Dillon: In the cases that I observed the brazing problems were mostly caused by poor joint preparation, too tight or too loose or misaligned. 02:59:13 James Dillon: As long as condensation is not constant, but intermittent for short periods of time, corrosion on the air side will be minimized and should be superficial. Stainless steel would be preferred for the support plates, but is much more expensive. Again, if contact time to condensate is minimized, air side corrosion of galvanized steel should be minimized. 03:00:57 David G..: Thanks for the fulfil answer, James Dillon. 03:01:37 James Dillon: Sacrificial anodes will only work when the material to be protected is submerged in the water along with the anode material. Typically condensate forms in thin films that won't produce such conditions. 03:05:00 James Dillon: It is common for air handler coils to be constructed with aluminum fins on copper tubes. If condensation is controlled and the environment isn't corrosive, there should not be substantial aluminum corrosion. However, if cooling water leaks the fins are very susceptible to corrosion. 03:32:10 Allison Boen: Is part of this workstream the idea to look outside OCP to the end users? 03:49:45 Mick Jones: Got to drop, thanks all for such informative talks! Definitely learnt several things.