<html><head><meta http-equiv="Content-Type" content="text/html; charset=us-ascii"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;">Hi all,<div class=""><br class="">I wanted to take a few moments to write some comments about DIRC bar box shipping, particularly the transit part, while the details are fresh in my mind. We (the shipping crew) should follow this up with a writeup, at least for internal consumption, if not some published document about issues with crate design. As this is somewhat from memory, some of the numbers might not be dead on -- it is semi-quantiative.<div class=""><br class=""></div><div class="">(Jerry: now you get a few of the gory details that I didn't quite have time to write while on the road.)</div><div class=""><br class=""></div><div class="">First, Yunjie deserves an award for keeping the crate and trailer telemetry working so that we had useable data at all the critical moments of the trip. We pushed the LORD wireless accelerometer system to the point of breakdown, which we realized was likely due to RF interference, and Yunjie was continually reconfiguring sensors or receivers to be sure we had the data we needed. We aren't sure if this is an inherent limitation on this product or a byproduct of the giant metal box that we put the system in. In any case, if there are future shipments to monitor, depending on the the number of desired sensors, a wired system would likely be more robust.<br class=""><br class="">Data: for archival purposes, we agreed that all accelerometer data as well as HD quality video used for monitoring the health of the bar boxes from both trips will be stored on the Indiana U. Scholarly Data Archive. This should provide access for an indefinite amount of time into the future should we realize a need to revaluate these aspects of the shipping process.<br class=""><br class="">Overall route and time: like last time we opted for a southern route that attempted to minimize time spent in mountainous regions. We traversed what remains of the Rockies on I-10. Max elevation is about 4800 ft which only slightly higher than that observed on I-5 just north of Los Angeles. We had 5 full days of driving. (Route map is attached with elevation profile.)<br class=""><br class="">The truck driver is limited to 11 hours drive time a day. This is a hard limit that is electronically regulated so in sparsely populated areas it requires "rounding down" to the closest stopping point that also has hotel facilities for the chase car crew. For the four nights on the road the truck driver stayed with the truck every night. Stopping points: Indio, CA; El Paso, TX; Mount Pleasant, TX; Gordonsville, TN. Total distance was about 3140 miles and we finished the 5th day with a few hours to spare on the trucker's driving clock.<div class=""><br class=""></div><div class="">Altitude: when crating it was noticed that bar box #1 exhibited a slight delamination of the Al-honeycomb layer near the window end of the box. The delaminated region was outlined with a pen so that we could see if it would grow when the differential pressure between the internal honeycomb and atmosphere increased at altitude. No noticeable growth in the delamination was observed when the box arrived at Jefferson Lab.<br class=""><br class="">Observations about shock and vibrations:<br class=""><br class="">During the external review for DIRC construction we proposed limits on allowable shock on the bar box during transit of 3.0 g in the up/down or left/right direction and 1.5 g in the fore/aft direction. These limits were based on static loading finite element analysis and some worst case scenario assumptions about inner construction and support of the bars. As an extra measure of safety we set limits that were at least an order of magnitude below what we thought was capable of causing damage in a static loading scenario.<br class=""><br class="">During the first trip last November, we loaded the single bar box crate in the middle of the truck. Throughout the transit we had one shock event in the up/down direction that was a bit over 1 g. We were pleased that we were well below our limits.<br class=""><br class="">Several things changed for this trip (some obvious in hindsight, some anticipated but unknown):<br class=""><br class="">We used a different style of double-face tape to attach accelerometers to the bar box. It was thinner and easier to remove after the shipment. This resulted in systematically higher readings. Tests with both types of tape post-shipment indicates that our previous measurements of shock to the bar box were likely underestimated by as much as a factor of 2-3 due to the double-face tape we used last time that had a thin foam core in it. The attenuation may be different in up/down (compressing tape) vs. left/right (shear stress on tape) which further complicates things.<br class=""><br class="">The loading arrangement for the crates was two forward and one in the back the trailer. The two forward crates experience similar shocks and ranges of shock; however, for many extreme events the shock on the exterior of the back crate was a factor of 2-3 larger.<br class=""><br class="">The air spring system on this trip used independent plumbing for horizontal and vertical which allowed us to soften the horizontal springs (which don't have to support the weight of the crate). This seemed to achieve the desired effect. Significant shocks in the left/right direction were much more rare on this trip.</div><div class=""><br class=""></div><div class="">Shocks experienced on the trip:<br class=""><br class="">As we finally got on I-5 and started south away from SLAC we had an event that registered over 4 g on the outer crate in the rear of the truck and the shock on the bar box just exceeded 3 g. This was due to resurfacing of I-5. If I am remembering the event correctly (there is some delay in the readout), the truck dropped off fresh asphalt onto unfinished road surface momentarily and then back up onto fresh asphalt (see notes below about timing of events). </div><div class=""><br class=""></div><div class="">During the first day we also noticed that shocks on the bar box in the rear of the truck at the level of 1 g, or what we thought at the time (and now know to be underestimated) was our maximum shock for the first shipment, seemed to be typical for the rear box in this shipment (one every 10 - 20 minutes). Nevertheless, these met our previously-agreed tolerance, and so we continued. The 3 g shock did give us pause and resulted in very careful examination of the video. After the first day we opted to reduce the air spring pressure in the vertical direction from about 23 psi to 19 psi. Due to the check valves in the system, some time is required (perhaps a day) for the true pressure to reduce on all springs. Due to systematically varying road conditions it is hard to compare quantitatively, but the frequency of >~ 1 g shocks on the back bar box seemed to decrease significantly after this air pressure adjustment. This adjustability -- thanks John Frye -- is a huge argument in favor of an air spring system, rather than fixed mechanical springs. </div><div class=""><br class=""></div><div class="">We managed to locate the same bump that gave us the maximum shock in the first shipment. This is just west of Memphis on I-40 leading up to the Mississippi crossing. There is a series of very short bridges on the floodplain that whose elevation is not aligned well with the road on either side. One of these generated a > 5 g shock on the outer crate, but the suspension system attenuated this to < 2 g, less than the shock experienced in the first hours of shipment. There was one more shock of just under 3 g on the rear bar box somewhere just north of the VA border on I-81. I think this was also associated with a bridge crossing.<br class=""><br class="">Future shipments (should they happen) may consider loading only the front of the trailer to be extremely risk averse, or adding more concrete ballast to the rear of the truck.<br class=""><br class="">Two lessons learned overall on shocks:<br class=""><br class="">It is very hard from the chase car to estimate the shock. Your eye is sensitive to the first derivative of position with respect to time, but for shocks it is the second derivative that matters. The amplitude of the trailer oscillation is rarely correlated with the shock. Some times we would see the whole trailer move up and down only to find the accelerometers barely detected the event. A good example that some of you may be familiar with is the stretch of I-64 under construction between Williamsburg and Newport News. The whole truck was rolling left and right and up and down. (The driver even commented on how rough it was.) But the accelerometers didn't indicate anything spectacular other than increased baseline vibration. On the other hand, from the car we barely noticed the bridge joint that generated the > 5 g shock: it is a single firm bump in the seat.<br class=""><br class="">Second, the time structure of the bumps is important and very hard to predict or develop an "all purpose" design to combat. Bridges in particular usually have a bump associated with getting on and off. Other things like large patches in road surface are similar. If the first bump gets either crate suspension or the trailer suspension oscillating then it is semi-random whether the second bump will be in phase or out of phase of the oscillation and thereby enhance or damp the shock to the box. We have several traces of acceleration with time where you can see dramatically different attenuation of shock on subsequent bumps.<br class=""><br class="">N2 purge:<br class=""><br class="">The three bar boxes in the second shipment were all much less leaky than the first. Prior to departure we plumbed each individually and measured the flow at fixed overpressure. Then we opened the exhaust slightly on two to try to equalize the flow rates. The combined flow rate for all three was similar to that of the box in the first shipment at the same overpressure. However, three times the volume of gas in boxes in this shipment meant that the flow needed for a fixed change in differential pressure when climbing or descending hills needed to be three times larger. This resulted in more dramatic swings of the overpressure when traveling than was observed during the first shipment. This caused problems in two specific places. Recall our specification was positive overpressure (clean N2 always going in) but keep overpressure less than 2" of H2O to avoid issues with the destructive overpressure relief valve on the bar box.<br class=""><br class="">When we descended on CA-152 near the San Luis Reservoir on our way over to I-5. This is a relatively short but fairly steep descent and it resulted in the overpressure going slightly negative for 1-2 minutes. After this descent we adjusted our "level ground" overpressure from about 0.5" H2O up to 0.8" H2O in order to give a little more margin from zero. This was beneficial when we descend I-5 north of LA which has a long 5% downhill grade. We managed to keep the pressure positive throughout this descent. (For reference, the atmospheric pressure changes about 14" of H2O per 1000 feet of elevation change: the bar box makes a great vertical speed indicator!)<br class=""><br class="">There were two instances in VA also coming out of the mountains that cause trouble with descents. There was a momentary approach to zero or slightly negative overpressure on a descent on I-81. The other more problematic one (which is visible on the route map) is a long straight (high speed limit) relatively steep descent on I-64 from where the highway crests the Blue Ridge Mountains. For this one, the overpressure went negative by probably a couple of tenths of inches of H2O on the scale of a few minutes.<br class=""><br class="">In these instances one has to do a global risk assessment: the speed limit in these stretches is 70 mph and traffic is going about 80 mph downhill. It is not safe to try to drive a car behind a semi-truck at 40-50 mph in one of only two travel lanes. We tried momentarily, but it is not pleasant to look in the rearview mirror and see a semi coming at you at 30-40 mph closing speed and a left passing lane that is full of cars passing you. (Imagine yourself as a very thin layer of salami in between two thick slices of bread). <br class=""><br class="">Finally, in one instance we lost flow of the purge and this failure was detected by the monitoring system and corrected. The regulator on the purge bottle had apparently vibrated closed (it was set at very low pressure). This was during a stretch of relatively flat road so exchange of gas with the environment from "breathing" going up and down hills was likely very small.<br class=""><br class="">We estimate that for the total trip a few hundred mL of air from the crate environment got sucked back into the bar boxes: not desirable but not devastating. This could be mitigated in the future by further opening up the exhaust lines and running at high flow and low overpressure. High flow requires larger pressures at the input to the filter bank used to purify the gas -- we would have been unable to go significantly higher without exceeding the mechanical capabilities of the pressure gauge on this part of the system. One could also run at higher overpressure, again we were limited by what we could do upstream of the filters. Also, the overpressure rises on ascents but there is usually always a "slow lane" to use on steep ascents on the interstate so the truck can be safely slowed. Too high overpressure is probably worse than a few moments of slightly negative pressure -- pick your poison. One may also consider equipping the truck driver with a variometer and asking him or her to keep the climb and descent rate within some limit as long as it is safe.<br class=""><br class="">Temperature/environment:<br class=""><br class="">Outside air temperatures peaked at 115 degrees F through Texas while in transit. The climate control of the refrigerated trailer worked remarkably well and the temperature was stable in the trailer at the level of a couple of degrees. Two things concerned us: hot air on the window of the back bar box when we opened the doors and dust. We therefore tried to do most work in the trailer with the doors closed in the harsh environment. Also the covers over the back bar box both protected it and provided some insulation from the immediate truck environment.<br class=""><br class="">Summary and comparison with our goals:<br class=""><br class="">* Two of the three bar boxes this shipment along with the box we shipped last November made the trip with no events exceeding our pre-defined tolerances for shock. The bar box in the back of the trailer for this shipment experienced to one event that slightly exceeded 3 g in the up/down direction on I-5 a few hours into the trip. No resulting damage was detected.<br class=""><br class="">* Challenges managing the N2 purge during elevation changes (particularly descents) resulted in total of several minutes of small negative pressure on boxes resulting in a flow from the environment into the three boxes on this shipment. The quantity of gas that flowed in is likely in the range of hundreds of mL. This may be comparable to, for example, breathing due to temperature variations while boxes are disconnected from purge for loading and unloading.</div><div class=""><br class=""></div><div class="">* Temperature was extremely well regulated inside the trailer even as the ambient temperature underwent variations at the level of > 50 degrees F throughout the entire trip.</div><div class=""><br class=""></div><div class="">* No adverse effects due to absolute altitude were observed.<br class=""><br class="">* Most importantly, in total, all four bar boxes arrived at JLab and show no visually apparent differences from when the were removed from the storage crate at SLAC.</div><div class=""><br class=""></div><div class="">Matt</div><div class=""><br class=""></div><div class=""><br class=""></div><div class=""><img apple-inline="yes" id="CE6EB2FA-FAA3-4995-A443-8F32500565AB" width="640" height="476" src="cid:851C1EE4-D886-4782-A47D-5378BF52EEF5@indiana.edu" class=""><br class=""><div class=""><br class="webkit-block-placeholder"></div><div class=""><br class="webkit-block-placeholder"></div><div class="">---------------------------------------------------------------------<br class="">Matthew Shepherd, Professor<br class="">Department of Physics, Indiana University, Swain West 330<br class="">727 East Third Street, Bloomington, IN 47405<br class=""><br class="">Office Phone: +1 812 856 5808</div><br class=""></div></div></div></body></html>