ASHRAE AN-04-2-3-2004 Shake Table Testing《振动台测试》.pdf

1、AN-04-2-3 Shake Table Testing James A. Carlson, P.E. Associate Member ASHRAE ABSTRACT Shake table testing is a test of equipment to determine whether the equipment can survive random oscillating motion. Equipment is bolted to a shake table that moves horizontally and possibly vertically in a prescri

2、bed motion, simulating an earthquake. Shake table testing onlyproves that equipment can survive the motion created by earthquakes. “Surviving an earthquake” can mean many diflerent things. Survival may mean that the equipment does notfly apart, that the equipment can start after the earthquake, or t

3、hat the equipment will oper- ate through the earthquake and keep running. The individual testing agency has historically subjectively applied the prescribed motion of a shake table based on interpretation of the requirements and shake table capabilities. This paper will provide an explanation of the

4、 requirements, criteria, and performance of the shake table test. INTRODUCTION There are few references available that describe shake table test requirements. IEEE-344 standard has been used universally to define the criteria for all shake table testing. This standard was initially published in 1975

5、 and last revised in 1987. It is still used today for many applications including test- ing mechanical and electrical equipment for nuclear power plants. This standard is very flexible and requires very well written specifications to meet customer expectations. A rela- tively new standard, IEEE-693,

6、 issued in 1997, covers the recommended practice for seismic design of power distribu- tion substations. In the case of nuclear power plants, these expectations are part of a license to operate. IEEE-344 stan- dard alone does not have sufficient criteria to perform testing without site-specific info

7、rmation. William Staehlin, P.E. In 2000 the International Conference of Building Offi- cials (ICBO) published “Acceptance Criteria for Seismic Qualification Testing of Nonstructural Components” (AC 156). This document defines the requirements for shake table testing from pre-test information to func

8、tional testing of the equipment using AC156. There is no discussion or commentary on requirements identified in the document. Extensive knowledge of shake table testing and application for equipment verification is necessary to apply AC156. This paper expands on the test requirements identified in A

9、C 156 with commentary for practical application. APPLICATION Code requirements are defined to test equipment and qualify the equipment will survive or will not degrade, causing additional damage. Inmost cases, the equipment selected to be tested that is the same equipment to be supplied to the job s

10、ite. However, the International Building Code (ICC 2000) and AC 156 do allow the manufacture to test and qualify a product line with one test, where the equipment tested is the largest size of the product line. It may be advantageous for manufac- tures to test the largest size of equipment in a prod

11、uct line at a maximum acceleration that may apply anywhere in the coun- try. The maximum acceleration may be larger than the specific application and the equipment may not be the same equipment that is manufactured for specific projects, but it may be cost- effective for future projects because of t

12、he cost to perform shake table tests. PRE-TEST INFORMATION Pre-test information includes a description of the unit to be tested. Unit weight, overall dimensions, and attachment configuration need to be provided to the vendor performing James A. Carlson is with Omaha Public Power District, Fort Calho

13、un, Nebraska. William Staehlin is with the Ofce of Statewide Health Planning and Development (OSHPD), Sacramento, Calif. 02004 ASHRAE. 339 the test. Rigid connections and vibration isolation are the two basic configurations to attach equipment to the building struc- ture. All attachment points shoul

14、d be defined in the specifica- tion or shown on vendor drawings and represent how the equipment is attached in its final location. Attachment points are vendor-specific for the equipment and, more importantly, for vibration isolators. The attachment to the shake table can use brackets and steel to s

15、teel bolts and may be different from the final installation where the equipment is usually attached to the building structure with concrete anchors. Anchor forces can be determined by using elements called “load cells,” placed between the component and the shake table. These load cells record tensio

16、n, compression, and, based on orientation, shear forces at anchor locations. The attachment ofthe compo- nent should represent the configuration as closely as possible. Vibration-isolated equipment adds complexity to the shake table test requirements. Multiple vendors are involved, providing equipme

17、nt and providing the vibration isolation devices. The building codes and AC156 are clear in the requirement for “system” qualification. The system, in this case, is defined as the component, the isolator(s), and possibly the attachment. The attachment can be accomplished by applying code formulas to

18、 obtain anchor forces. Normally the vendor providing the vibration isolation is responsible for the attachment to the equipment and to the building structure. Complete attachment details should be provided with the vibration isolation for the shake table test as well as the final attachment requirem

19、ents. Other information required before the shake table test can be implemented includes functional acceptance criteria and the minimum earthquake shaking requirements in the form of a required response spectrum (RRS). Functional acceptance defines the survivability of the equipment as described in

20、the abstract. AC1 56 differentiates the functional requirements using the importance factor. If the importance factor is 1.0, then the shake table test is only to ensure the structural integ- rity of the equipment. That is to say, the equipment will not move from the attachment points and any intern

21、al parts or assemblies will not separate physically. Permanent deforma- tion and yield of steel members is allowed. This is only a mini- mum set of requirements and may not represent the building owners requirements. If the importance factor is greater than 1 .O (or 1.5), then the equipment is deeme

22、d to be essential for continued opera- tion of the facility andor essential to maintain the operation of the critical life-support systems andor contains hazardous materials (quantities defined by code) harmful to humans or the environment. This type of equipment is required to be proven operational

23、 before and after the shake table test. It is important to clearly identify the functional requirements and describe how the functional requirements are proven. There are many ways to prove operational. For fans and refrigeration systems, the minimum cooling load or airflow requirements should be id

24、entified. Do not use nominal values because there may be some degradation of the equipment from the testing, shipping, andor manufacturing. However, proving equipment is operational requires power supplies to turn motors and possibly load banks (coils if necessary) to show that the performance of th

25、e equipment meets the functional requirements. These functional requirements should be demonstrated prior to the shake table test so that an accurate evaluation after the shake table test can be completed. Using fans as an exam- ple, the fan should be started and measurements of operating parameters

26、 recorded. Fan airflow is not really required to show operability. Structural integrity, fan speed, and bearing temperature can be used to define operability. For example, the fan housing shall not deform, causing the fan blade to rub against the housing after the test. The fan speed may be spec- if

27、ied to be within 2% of the original fan speed after the test if the airflow is acceptable within 2% percent. Acceptable airflow range should be specified. The bearing temperature may be within 50F of the original after the test. Another example may be a freezer for blood storage in a hospital. The f

28、unctional requirement is to keep the blood frozen. In addition to the operational requirements of the refrigeration system, the door and structural assembly must stay intact to keep the cold air in the container. In this case minimum door displacement may be provided whereas the latching mechanism m

29、ay or may not be required to work. Each and every equipment and application may have very different functional requirements. To prove operation requires specific criteria. These operational tests must be defined at the begin- ning of the project and included in the contract with the shake table test

30、 vendor. Most equipment tested is also the actual equipment provided in the field and installed. Rarely is there a test piece of equipment provided only for testing and later used for spare parts. This is sometimes done at the request of the building owner for equipment that has a very high seismic

31、required response spectrum. Since this test is expensive and the equip- ment will be used, most owners would like some operational testing performed after the test to ensure that the equipment will operate when installation is complete. In addition to the functional requirements, any restric- tions

32、or limitations of the equipment should be included in the testing requirement section. This includes hazardous materials handling (oil in transformers and hydraulic systems) and/or any special instructions. Sometimes a factory representative is required to be present during the testing and provide a

33、ssistance in removing shipping restraints. All shipping restraints should be removed or relieved for the test to be valid. Digital controls may be secured with additional clips. There are a few cases where control boards are mounted on springs. These are always a problem. REQUIRED RESPONSE SPECTRUM

34、Information for performing acceptable shake table tests includes an acceleration input that represents the earthquake. Acceleration input is usually identified in the shake table spec- 340 ASH RAE Transactions: Symposia -1 Horizontal Frequency (Hz) Figure 1 Required response spectrum. ification in t

35、he form of a floor/ground spectrum. Occasionally there is a site-specific ground response spectrum. If the equip- ment is mounted on elevated floors or on the roof of the facil- ity, then the site-specific ground response spectrum must be adjusted to represent the actual floor response spectrum wher

36、e the equipment is installed. Current criteria have provided a prescriptive approach with a prescribed required response spectrum (RRS). The prescriptive approach in AC156 uses a normalized RRS with a maximum value provided by equations equal to the equivalent horizontal force factor Fp used in the

37、Znterna- tionul Building Code (IBC), 2003, and ASCE 7-02. Figure 1 shows the generic horizontal and vertical input spectrum that are acceptable to qualifj most equipment. A, is the peak of the curve. From AC156, A, is Fp assuming a flexible component normalized for the equipment weight. A, is A, nor

38、malized for the in-structure equipment amplification factor up= 1 .O, which represents rigid components with a fundamen- tal period less than or equal to 0.06 seconds (2 16.67 Hz). All equipment is assumed to have a 5% damping factor. This damping factor accounts for small plastic type deformations

39、experienced by equipment. The vendor performing the shake table test will use this spectrum to develop the test input values. The RRS contains two curves as indicated above. The horizontal motion response spectrum is the acceleration including the elevation of the equipment installed in the build- i

40、ng relative to the roof level. The vertical motion response spectrum is simply two-thirds of the horizontal spectrum. AC156 requires that the horizontal and vertical be simulta- neous and phase incoherent inputs during the shake table test. SHAKE TABLE TEST-SINE SWEEP The first part of the test as r

41、equired in AC156 is a sine sweep. Sine sweep is a method to determine the resonant frequencies ofthe equipment. This test is a basic test where the shake table isolates back and forth at 1 Hz and increases to 33 Hz. Amplitude of the isolation is approximately 0.5 in. to represent a small acceleratio

42、n in the amount of O. 1 2 0.05 g at 1 Hz (Le., one g is equivalent to 32.2 f/sec/sec or approxi- mately 386 in./sec/sec). As the frequency increases, the ampli- tude must decrease at a factor equivalent to the increase in frequency squared to maintain the same acceleration. In simple terms, the ampl

43、itude will decrease by a factor of 100 when the frequency increases by a factor of 10. So, at 10 Hz, the amplitude will be about 0.005 in. At 33 Hz, one cannot see the table vibrate back and forth, but the noise developed sounds like a jet engine. When shake table testing was first introduced, some

44、oper- ators did not understand the requirement to reduce the ampli- tude at higher test frequencies. So as the operator increased the speed of the shake table, maintaining the same amplitude, the resultant acceleration went from a small value to 20 to 100 g?s. At these high accelerations, the equipm

45、ent and sometimes the shake table was severely damaged. Now computers are used to control the shake table testing, eliminating unfortunate acci- dents. This sine sweep is performed to determine the natural frequency of the equipment. Accelerometers are attached to the equipment and measure the respo

46、nse from many locations. As the equipment responds to shaking, certain frequencies have a higher response or plastic deformation and movement of equipment. For equipment with metal cabinets or enclo- sures, this natural frequency may be from about 12 Hz to as high as 20 Hz. According to the Internat

47、ional Building Code, (ICC 2000) any equipment with natural frequencies above 16.6 Hz is considered rigid and will not have any additional response or higher acceleration than the shake table will develop. This is a basic definition for a single degree of free- dom system. Most earthquakes have the h

48、ighest amplitude of acceleration and peaks at around 5 Hz. Usually there is another smaller peak around 12 Hz. Sine sweep really only applies to rigidly mounted equip- ment. If the equipment is mounted on spring isolators, sine sweep testing only finds the natural frequency of the equip- ment weight

49、 and spring constant (k) combination. The natural frequency of the equipment cannot be determined. The code does not make any allowances for performing the sine sweep based on the type of attachment. AC156 requires the sine sweep to be performed. A side note for vibration-isolated equipment is that the spring constants of the spring are selected based on the trans- missibility of noise and vibration from the equipment to the building structure. Typical selection also sets up a natural frequency ofthe equipment weight and spring constant around 2 to 5 Hz, which corresponds to the peak of t