A comparison of C0 and C2 monitoring in renal transplant patients Bülent Yi¤it*, Melih Kara, ‹brahim Berber, Gürkan Tellio¤lu, Funda Türkmen, ‹zzet Titiz Haydarpaﬂa Numune Research and Educational Hospital, Renal Transplantation Unit, ‹stanbul, Turkey Objectives: To compare the effects of Cyclosporine CsA-0 (C0) and CsA-2 (C2) level monitoringon acute rejection (AR) and infection rates, toxicity and graft survival in patients with renal trans-plantation.
Getting It Right
CFD Simulation saves Figure 1: Raised-floor simulation
Mentor Graphics with superimposed CFD grid
using their own tool to design two new Figure 2: Dropped-ceiling proposal showing non-uniform pressure
EING a leader in the electronic
and temperature distribution in a room with 10-ft ceiling height
design automation (EDA) market
Bhas its upside when you want to
build not one but two centralized
datacenters. With a company full of talented
engineers and their own highly revered
Computational Fluid Dynamics (CFD)
modeling tool, FloVENT™, it is no wonder the
company decided on a purpose built center
as opposed to outsourcing to a third party.
The Mentor design team were commissioned
to design an effective cooling architecture
and then form the configuration of the
building around that. While there were some
basic guidelines to define the size and shape
of the structure, the details of ducting,
venting, and internal floor plans were driven
by the servers' needs.
The company's primary goal was to save power by minimizing the use of air conditioning to cool their datacenters. With one of the new centers based in Oregon and another in Shannon, Ireland, Mentor Graphics set out to consolidate the resources of more than 20 local centers down to just two. The decision to build custom centers was driven by the steady, costly growth in the company's overall server heat load, which has risen by 33% annually. The climate of both Wilsonville and Shannon helps to keep the datacenters from overheating - Mentor predict they will only have the need to use air conditioning 10% of the time. 22 mentor
CFD Solutions at Your Fingertips
To begin, the simulation was gridded into
many thousands of very smal cel s that
could be individual y analysed and integrated
into a composite flow-thermal view. Figure
1 shows such a grid superimposed onto a
temperature slice map of a raised-floor HVAC
concept. This variable cel sizing provides
higher resolution in specific areas such as the
server rack interiors.
Historically designing a building's air and thermal flow involved formulating estimates and judgments based on experience, and then submitting the design to an external specialist for CFD analysis. This approach can deliver the necessary accuracy but is time consuming, often taking weeks for each round of evaluation. Nowadays organizations are bringing CFD in-house, using design teams to perform flow analysis to rapidly model design iterations. Mentor used their own CFD product, FloVENT, in the design of their centers and were able to generate reliable thermal and flow data that guided engineers in proof-testing new ideas.
When considering their datacenters, the
Mentor team contemplated four methods;
the traditional raised floor, a dropped or
suspended ceiling, and two further methods
that involved hot air confinement. The
decision about the basic air circulation
approach for the server rooms il ustrates the
benefit of performing flow analysis local y.
A simple raised floor design proved inadequate after CFD studies predicted pressure problems under the floor. The raised floor should function as a plenum that regulates pressure and distributes air efficiently. However the volume beneath the floor tiles was insufficient and therefore incapable of acting as a plenum to ventilate the air flow as the air would not reach all the server racks uniformly. To correct this, it would have been necessary to increase the floor's height considerably, adding to an already costly project by as much as 28%.
When examining the suspended ceiling design, the configuration involved hot air from the servers traveling through a chimney system to a plenum space formed by the ceiling itself, and onward to rooftop coolers. The capacity of the plenum was again a problem. To design a plenum of mentor.com/mechanical
Figure 3: Initial layout of the chimney/
Figure 4: A centered return plenum ensured
consistent pressures throughout the collector duct
"Using FloVENT®, a powerful CFD software tool, has been a real bonus for the datacenter design team. After very little training the design team has been able to investigate various permutations virtual y, to manage airflow and temperature in the datacenter. We were able to experiment with different layouts and technologies before we finalized the optimized design. Indeed, this process has saved us time and money and it should lead to very significant energy savings over the years the data centers wil be in service." Ananthan Thandri, Chief Information Officer, Mentor Graphics
Aerial view of the Wilsonville datacenter under construction
sufficient capacity would constrain the aisles. These flaws are a result of uneven design was abandoned after CFD analysis
ceiling height in the server room, and the pressures throughout the room.
revealed that the aisle was too hot at
remaining room volume wouldn't allow 120° F, these temperatures would impact
the pressurized incoming air to distribute Capturing Hot Air
a technician's ability to perform routine
evenly. Figure 2 compares a CFD pressure The next two methods considered involved
maintenance on the servers.
"slice" with an equivalent temperature the confinement of the hot air leaving the
contour slice, both spanning the room at server racks and redirecting it. This "hot
Figure 3 demonstrates an alternative the 6 ft. height of the server racks. Problem aisle containment" technique forces and
design incorporating a chimney system that areas in aisles one, three, five and seven confines the air in an entire aisle between
connected chimneys directly to individual were instantly revealed by the simulation, two rows of racks and directs it toward an
server racks, drawing heated air from them showing areas of high heat in the "cold" exhaust duct. Early on in the process, this
and guiding it to a collector duct and then 24 mentor
Ananthan Thandri, Chief Information
Officer, Mentor Graphics
into a return plenum. Using CFD simulation, a range of scenarios was tested for the chimney/collector including a worst-case example in which the rear doors of several server racks were open for maintenance at the same time. The results confirmed first that the aisle temperatures were much lower at approximately 80° F making it safer for engineers, and secondly that server racks still supplied at least 66% of their hot exhaust air into the chimney when their doors were left open. Both findings supported the efficacy of the chimney/collector architecture. Final y, a number of "what-if" simulation scenarios aimed at refining individual components were run. They proved helpful to the Mentor team when evaluating the chimney/col ector design. The CFD plots in Figure 4 depict two alternatives for the return plenum layout. As a result of the simulations minor changes to the design were identified that could improve air flow in the duct. Scenario "B" depicts the design changes that resulted from conclusions in Scenario "A", namely the positioning of the plenum in order to relieve the pressure on the servers nearest to it. Further CFD experiments produced a duct design with angled corners that minimized this disturbance.
The Mentor Graphics datacenter project is still underway in Wilsonville, with completion expected in December. Even as the construction progresses the Mentor team are refining HVAC control algorithms based on further CFD studies and real-time data from the already completed Shannon datacenter. With the aid of FloVENT, a CFD tool for HVAC and datacenters, major design issues were investigated and improved. As well as this the project team were awarded a grant from the Oregon Energy Trust for meeting stringent energy compliance targets. CFD simulations have proved invaluable to expose design flaws and dictate the correct design path before any concrete has been poured, and saving time and ultimately money by getting it right the first time.
Guidelines for pediatric advanced life support Official reprint from UpToDate® www.uptodate.com ©2014 UpToDate® Guidelines for pediatric advanced life support Section Editor Deputy Editor Eric Fleegler, MD, MPH Susan B Torrey, MD James F Wiley, II, MD, MPH Monica Kleinman, MD All topics are updated as new evidence becomes available and our peer review process is complete. Literature review current through: May 2014. This topic last updated: Jan 31, 2013.