Although the topic of sound and vibration control within air conditioning is quite complex, we describe here some of the fundamental influences (Schaffer, 1991):

• Fan rotation speed is a major factor in sound propagation
  dB 1 = dB 2 + 50log 10 (rpm 1/rpm 2)
  Thus, a fan moving at a 20% slower speed should exhibit a 5 dB drop in sound level
  
• Other factors:
  - Vortex shedding: turbulent eddies in the wake of the fan blade tips
  - Turbulence due to the obstructions in the intake or exhaust wake
  - Fan motor vibration
  - Harmonic resonance associated with the number of blades
  - Fan interaction with compressor noise

Within our research, a key emphasis to reduce sound levels was to operate fans more slowly with efforts made to suppress fan tip vortex shedding and harmonic resonance associated with fan blades.

Baseline Air Conditioning Unit and Test Facility

For our testing, we used a standard 3-ton (10.6 kW) air conditioning system produced by a major U.S. manufacturer. The system uses R-22 refrigerant, although our evaluation was done with condenser fan only operation. The system has a rated SEER of 12 Btu/Wh (SCOP = 3.51 W/Wh) when mated with a compatible evaporator and air handler. The 19" (48 cm) fan in the original outdoor unit consists of four metal paddle blades, powered by a six-pole 1/8th hp (0.09 kW) PSC motor with a rated flow of 2400 cfm (1,133 L/s) for the condenser. As measured in the baseline condition, the fan motor drew 190-197 Watts at 208 Volts and produced 2,180-2,200 cfm (1,029 - 1,038 L/s) turning a 1,010 rpm. At the 95/80/67 ARI test condition the fan power for the entire AC system represents about 6% of total system power.

For the research, we needed to accurately measure power, condenser air flow, fan motor power and rpm as well as environmental conditions. Secondly, we desired to measure sound levels. For diagnostic purposes, we also used flow visualization tools (smoke pencils and flow wands) to aid our understanding of the air flow dynamics.

An indoor test facility was constructed. A precision power transducer provided motor power measurements ( +1 Watt resolution) and air temperature and relative humidity. As the facility had 208 single-phase power, this electrical source was used for the measurements involved. A laser tachometer was used to measure fan rpm and a precision portable dB meter was used to measure nearby sound levels. A digital manometer was used to measure static pressure within the condenser underneath the exhaust fan. As condenser air flow was a critical measurement, we constructed flow measurement chamber in conformance with ASHRAE Standard 51-1985. The constructed outlet duct chamber with flow straighteners and settling screens was then calibrated at another facility with NIST traceable air flow equipment. The final chamber was estimated to yield an absolute air flow measurement accuracy of approximately + 5% (125 cfm or 59 L/s). The relative air flow measurement accuracy was much better. We found the equipment could reliably measure changes in air flow as small as 20 cfm (9 L/s) out of a 2,500 cfm (1,180 L/s) air flow.

Development of New Fans

In designing fans, our objective was to create the most efficient designs while operating at low rotational speed to reduce fan noise. We also looked to create robust characteristics which would provide good performance over a range of static pressure. This is important as pressure rise can change as condensers foul or due to heat pump outdoor unit frosting during winter operation. Generally, the true airfoils we used have flatter fan curves than those for curved metal bladed fans.

Over a period of two years, a total of five different fans were designed and built (designated A-E) with a series of sub-variations on each. The differing fan configurations were targeted for differing rpm ranges, static pressure rise and sound characteristics. The Original Equipment Manufacturer (OEM) design was a stamped 3-bladed metal fan. Fans A and D consisted of three equally spaced blades, with tapered and twisted air foild. Fan D was designed for a higher pressure rise. Fan A5 was an asymmetrical 5-bladed design. Fan E had forward curved blades, intended to assist with sound reduction. Each of the designs were evaluated by computer simulation and then produced as three dimensional coordinate files that could be used to describe the complex shapes. Rapid prototyping was used to physically produce the fan blades. Each fan was then hand-mounted onto a produced hub and speed balanced before evaluation on the test stand.

Figure 5. 19" Fan A5 with asymmetrical blade spacing.

The fan design with an asymmetrical alignment has unequally spaced blades. This configuration was explored to potentially lower noise levels. This technology has been previously developed for helicopter rotors (Kernstock, 1999), but not previously utilized for AC condenser fans. The sound of air rushing through an evenly spaced fan rotor creates a resonance frequency with the compressor’s hum, leading to a loud drone. But if the blades are not equally spaced, this resonance is reduced producing lower ambient sound levels. With our invention, we took advantage of the asymmetrical characteristics using a five-bladed fan design where the
fan blades are centered unevenly around the rotating motor hub (Figure 5).

We tested each fan design with 1/8 hp (0.09 kW) PSC motors either with six poles rotating at approximately 1075 rpm or eight poles rotating at 850 rpm. Over one hundred tests were conducted over an 18 month period.

Advanced Diffuser Design

Diffusers are an expanding duct which provides recovery of air static pressure by reduction of the flow velocity as the flowing air mass expands. Practically, the condenser fan air velocity is lowered prior to exhaust, thereby increasing the overall mass flow rate from the system. The exhaust configuration of a standard unitary air conditioner consists of a short 4" (10 cm), 10 degree divergent diffuser covered by a slotted grate or wire grill with the fan is nestled in the bottom of the diffuser section.

Examining interactions between high efficiency propeller designs and external static pressure, we determined that an optimized diffuser section would allow large improvements in air moving efficiency. Diffuser theory would suggest that large improvements in fan efficiency are possible by lengthening the diffuser stage (Blevins, 1984). Theoretically, an 18"(46 cm) diffuser should provide about 25% added pressure recovery over that from a short 4" (10 cm) diffuser (Japiske and Baines, 1993). While a longer length would provide still greater pressure recovery, we judged an 18" (46 cm) height to be the maximum practical for consumer acceptance.

Thus, we constructed a larger 18" (46 cm) tall 7o divergent diffuser with the motor and fan located in the bottom of the assembly. Figure 6 shows the overall assembly as produced with the elongated diffuser. Essentially this modifies the overall fan design from more of a shallow ducted propeller to a true tube-axial design.

Figure 6. Diagram of the improved condenser fan with enhanced diffuser.

The diffuser increases the exhaust diameter from 19.75" (50 cm) to 24" (61 cm) at the wire grill top. While testing with an experimental stator stage did not demonstrate any added flow, we did find other changes in geometry to yield modest improvement. As the motor occupies the center of the diffuser, the swirl set up by the fan as it expands through the diffuser tends to collapse on the low pressure zone immediately behind the motor. Through trial and error, we found that by using a smooth conical center body on the other side of the motor, we could increase flow by 20 cfm (9 L/s) and reduce power by about 2-5 Watts.

Reducing Tip Clearances and Sound Control

The functionality of an air conditioner condenser exhaust is essentially analogous to a ducted fan. Research done over the last twenty years within aeronautical engineering has shown that tip clearance of ducted fan blades to diffuser walls is critical to performance (Rajagopalan and Zhang, 1989; Abrego and Bulaga, 2002).

Unfortunately, low tip clearances are practically difficult in manufacture due to required tolerances. Should fan blades strike a solid diffuser wall, the fan blades or motor may be damaged or unacceptable noise created. Thus, in air conditioner fan manufacturing, the fan blades typically have a gap of 0.3 to 0.4 inches (0.8 - 1.0 cm) to the steel sidewall diffuser. This large tip clearance has a disadvantageous impact on the ducted fan’s performance.

Considering the desirability of low sound levels for AC condensers we examined interesting work done at NASA Langley Research Center showing how porous tipped fan blades in jet turbofan engines can provide sound control by reducing vortex shedding – a known factor in the propagation of excessive fan noise (Khorrami et al. 2002).

Based on the research, we postulated that rather than porous fan tip, a porous diffuser sidewall could achieve the same result. This was done by obtaining commercially available 3/16" (0.5 cm) open cell plastic foam 1 ½" (3.8 cm) wide, and applying it to the inner wall of the diffuser assembly swept by the fan blades. In actual application a UV stabilized open cell neoprene foam would likely be used. As shown in Figure 6, the foam is applied within the diffuser assembly over the swept blade region to breakup fan blade tip vortices and reduce sound. We also used a solid tip clearance strip to test the differences. Whereas the solid strip actually increased fan noise, the open cell foam strip reduced noise markedly (see Table 1).

One added advantage was that the foam can be used to produce very close tip clearances in ducted fans with no danger to the moving blades. Any contact with the foam inner liner will be quickly worn away to yield ideal fan tip clearances. This was verified in overnight tests where tolerances were exceeded. A final advantage is simplicity and cost effectiveness. This is a simple change that can potentially produce large improvements in acoustic and air moving performance.

We estimated the flow and sound impacts of the invention by carefully measuring performance of two fans. Sound levels from fan only operation were measured using hand-held dB meters at the prescribed distance used for ARI 270-1995 for the horizontal measurements. The results in the Table 1 show a large improvement in airflow as well as sound advantages.

Table 1 Impact on Performance of Reduced Tip Clearance Using Foam Sound Control Strips
Case	Flow	Power	Normalized CFM/W	dBA
OEM Fan with slotted grill (1000 rpm) with standard diffuser and top
Original Configuration	2200 cfm	190 W	11.6	63.0
A5 Fan with 8_pole motor (850 rpm) with extended conical diffuser
As is (~1/4" clearance) Tip clearance <1/32" foam	2110 cfm 2300 cfm	135 W 141 W	15.6 16.3	62.0 60.0
A Fan with 6 pole motor (1100 rpm) with conical diffuser
As is (~1/4" clearance) Tip clearance <1/32" foam	2400 cfm 2610 cfm	139 W 145 W	17.3 18.0	64.5 61.0

Note the improvements in air moving efficiency. As the shaft power requirement increases between the square and the cube of the flow quantity, the advantage of the foam strip for A5 (2,110 to 2,300 cfm) represents a measured improvement in the air moving efficiency of nearly 23%. Moreover, at three feet (1m) away from the condenser, we measured sound reductions of at least approximately two decibels (15% more quiet to the human ear). In contrast, we had previously attempted a number of other suggested improvements (forward swept blades, dimpled air foils and winglets) which did not produce any measurable sound reduction.

Tests Results

The fans were evaluated with the standard slotted grill top and later with the improved diffuser configuration (elongated diffuser with foam tip clearance strip and conical center body). We also tested with differing PSC motors (six vs. eight pole operating at 1075 and 850 rpm, respectively). The process involved measuring performance and then evaluating the changes in a comparative manner to isolate the best options. The best performing configurations with and without the enhanced exhaust are summarized below. Table 2 provides numeric data and Figure 7 shows the graphical results.

Table 2 Comparative Performance of Fans & Diffuser Elements
	Top	Fan	Motor	Flow (cfm)	Power (W)	CFM/W	Sound (dBA)
*	Slotted Slotted Slotted	OEM D A5	6-pole 6-pole 8-pole	2200 2190 1660	190 150 130	11.2 14.6 12.8	63.0 65.0 62.0
**	Diffuser Diffuser Diffuser Diffuser Diffuser Diffuser Diffuser	OEM A5 A A E E D	6-pole 8-pole 8-pole 6-pole 6-pole 8-pole 6-pole	2250 2300 1930 2610 2500 1825 2590	173 141 111 145 132 109 150	13.0 16.3 17.4 18.0 18.9 16.7 17.3	63.0 59.0 58.0 66.0 65.0 61.0 66.0
***	Wire-foam Wire-foam	OEM A5	6-pole 8-pole	2250 2110	188 146	12.0 14.5	62.0 60.0
* Standard configuration, metal bladed fan
** Preferred configuration, with advanced fan and diffuser with foam tip clearance strip
*** Preferred configuration with standard top, wire grill and foam control strip

Figure 7. Performance of different fans with and without the enhanced diffuser against the original configuration

Results are shown for each fan, both with the slower 8-pole motor (850 rpm) and the faster 6-pole motor (~1075 rpm). The best performing fan with the standard slotted grill top and short diffuser was the three-bladed Fan D which was designed for a higher pressure rise. It produces the same flow as the standard OEM fan (~2200 cfm or 1,038 L/s) with a power consumption of 150 W vs. the 190 Watts for the standard fan at 208 Volts. Sound levels are similar to the standard configuration.

We also did a test of the original configuration with a wire grill rather than slotted top above the short diffuser. The wire grill showed superior performance– likely due to a lower pressure rise above the fan. The power of the original metal-bladed fan dropped by 2 Watts with equivalent flow. When the foam tip clearance strip was added to the original short diffuser, flow increased by 50 cfm (24 L/s) along with a small drop in sound level.

Also shown are tests using the enhanced diffuser with Fan A, Fan A5 (a 5-bladed asymmetrical version of Fan A) and Fan E. Two fans (B and C) are not shown as results were not promising.

Note that the A5 fan with the diffuser improves air moving efficiency (CFM/W or L/s/W) by greater than 46%. With the enhanced diffuser, this was our preferred configuration. It produced 100 cfm (47 L/s) more flow than the standard configuration while still reducing power by 49 Watts. Unlike Fan D, it also reduces sound since the enhanced diffuser allows the use of a slower turning fan with an 8-pole motor to achieve better flow. Fans A and E allow superior flow over the original configuration at even lower power, although sound levels are increased. These results suggest that a fan such as A5 could be run at even lower RPM using a variable speed motor to provide greater power and sound reductions or conversely run at higher RPM to provide greater flows.

Tests with a Variable Speed Motor

To examine how the new fan and diffuser designs would compare with the standard design at a given flow, we obtained a brushless direct current (BDC) 1/3 hp (0.25 kW) motor from a leading U.S. manufacturer. It was programmed to be variable in speed from 0 - 1200 rpm in response to a pulse width modulated DC signal. These motors tend to be more efficient than PSC motors– by approximately 15% at full speed, but with large differences at lower speeds. Our tests verified these expectations and also showed the full advantages of the air foil fan designs as well as the enhanced diffuser when flow was equivalent to the OEM design.

For instance, the OEM fan with the standard top and 1/8 hp (0.09 kW) PSC motor required 190 Watts to provide 2,200 cfm (1,038 L/s) of flow at 208 Volts. The original fan with the enhanced diffuser and the BDC fan required 125 Watts to provide the same flow (2200 cfm). Thus, the diffuser and the BDC motor produced a power savings of 65 Watts or 34%. However, Fan A5 with the enhanced diffuser, and BDC motor, only required 86 Watts to provide the same flow – a reduction in power of 102 Watts or 55%. The same tests also showed that by increasing rpm with the A5 fan up to 1,060 rpm, flow was increased to 2,410 cfm (1,137 L/s) with 106 W power draw.

Within our research, we established that modulating outdoor unit fans speeds may have attractive performance and sound tradeoffs. We speculate that much of the nuisance of air conditioner noise comes during nighttime when ambient sound levels are low and occupants are asleep. This suggests that the BDC condenser fan speed might be modulated to high speed during very hot daytime periods. For instance, fan speed could be set to high above 94oF (34oC) while a low speed would be used when the outdoor temperature wa less than 84oF (29oC). This would substantially reduce fan noise during evening hours while preserving best peak performance during hot afternoons. Figure 8 below shows the comparative performance of the OEM configuration, the OEM fan used with the enhanced diffuser, and BDC motor with the improved fan design.

Figure 8. Impact of BDC motor and improved fan blades on condenser performance.

In Figure 8, the impact of the diffuser and BDC motor can be seen as well as the influence of the improved fan blades. Finally, Figure 9 below isolates the airflow improvement created by the tip clearance and sound control foam strip when evaluated at differing flow points with the BDC motor and the high performance fan.

Figure 9. Impact on performance of Fan A5 used with elongated diffuser with and without tip clearance foam strip.

Additional Laboratory Measurements

In May 2004, we took the prototype condenser top to a major U.S. air conditioning manufacturer who possessed a sound room and air flow measuring facilities to allow verification of measurements. Within their laboratories, we found that measured sound reductions of the new diffuser top were only 1-2 dB as measured according to the ARI 270-1995 standard. This is likely due to the tendency of the longer diffuser to broadcast sound upward to the overhead microphones. We also were able to verify the flow and power reductions previously measured. Power consumption of the PSC motors, when tested at 230 volts was about 10 watts more than test done at 208 volts. However, the savings with PSC motors at 230 volts could be made similar to what we measured by slightly reducing the shaft wattage of the motors for the improved blades. This would reduce power waste as the motors approach synchronous speed at higher voltage.

Conclusions

A research project designed, fabricated and tested high efficiency air conditioner condenser fans to improve flow characteristics while dropping motor power. We also examined potential changes to the condenser exhaust configuration to enhance performance. The primary objective was to reduce motor power while providing similar or superior flow to the standard configuration. A secondary objective was to provide sound reductions which is important to consumers.

Within the effort, we developed new twisted and tapered propeller air foils that demonstrated greater air moving efficiency. Fan only savings were 40 watts (21%) for the same motor and condenser top.

We also showed how a lengthened diffuser with a conical insert after the motor can improve air moving efficiency by over 16% for standard fans and over 27% for high performance fans. Fan tip losses and associated vortex shedding was reduced though the use of a porous foam strip to improve air flow performance while helping to reduce sound. This also allows slower fans to provide superior air moving performance along with sound control. On the negative side, however such design would lengthen condenser height by approximately 18" (46 cm) and somewhatincrease costs by increasing sheet metal requirements.

The fundamental project achievements are summarized below:

• Provides 49 Watt reduction in fan power (141 W vs. 190 Watts) with PSC motors at 208 volts.
  
• Increases condenser air flow by 100 cfm or 47 L/s (5% increase in fan flow).
  
• Provides 102 W power reduction with BDC motor.
  
• Reduces fan-only ambient sound level by 1-2 dBA. Ground level sound reduction is greater.
  
• BDC motor allows lower fan speeds for ultra-quiet night operation, higher flows for maximum capacity during very hot periods (temperature based control)

Key Technologies

• High efficiency 5-bladed asymmetrical fan moves air quietly at lower fan speeds.
  
• Diffuser top for effective pressure recovery allowing increased air flow at low speeds.
  
• Conical center body reduces losses in exhaust swirl.
  
• Foam sound control strip to reduce tip losses and fan tip vortex shedding.
  
• Patents pending: U.S. Application Serial No. 10/400,888, Provisional applications 60/369,050 / 60/438,035 & UCF-449CIP; WhisperGuard (UCF-Doc

Acknowledgments

This work was funded by the U.S. Department of Energy within its Building Technologies Division. Thanks to Terry Logee for his support. Gary Nelson and Ron Rothmann with The Energy Conservatory assisted with development and calibration of the air flow measurement equipment. At AeroVironment, Inc. John Gongola and Guan Su assisted with creation of the computer generated designs. In particular, we appreciate the great skill of Shep Shepperd of Merritt Island, FL in provided the precision machining to assemble the prototypes and exhaust manifold configurations.

References

A. I. Abrego and R. W. Bulaga, “Performance Study of a Ducted Fan System,” NASA Ames Research Center, American Helicopter Society Aerodynamics, Acoustics and Test Evaluation Technical Specialists Meeting,” San Francisco, CA, January 23_25, 2002.

ANSI/ASHRAE 51-1985 (Revision: ANSI/ASHRAE 51/AMCA 210_1999): Laboratory Methods of Testing Fans for Rating, American Society of Heating, Refrigerating and Air Conditioning Engineers, Atlanta, GA.

ARI, 1995. ARI 270-1995: Sound Rating of Outdoor Unitary Air Conditioning Equipment, Air Conditioning and Refrigeration Institute, Arlington, VA.

ASHRAE, 2004, “Chapter 18: Fans,” HVAC Systems and Equipment, American Society of Heating, Refrigerating and Air Conditioning Engineers, Atlanta, GA.

R. D. Blevins, Applied Fluid Dynamics Handbook, “Nozzles, Venturis and Diffusers,” Van Nostrand Reinhold, NY, 1984.

J. S. Bradley, 1993, “Noise from Air Conditioners,” Institute for Research in Construction, National Research Council of Canada, Ottawa, ONT.

DOE/EIA, 1999. A Look at Residential Energy Consumption in 1997, Energy InformationAdministration, DOE/EIA-0632 (97), Washington, DC.

S. K. Fischer and C. K. Rice, 1983. The Oak Ridge Heat Pump Models: I. A Steady_State Computer Design Model for Air_to_Air Heat Pumps, ORNL/CON_80/R1, Oak Ridge National Laboratories, August, 1983

R. H. Green and L. Roberts, 1996. “The Effect of Air-Coil Design on the Performance of Heat Pumps and Air Conditioners,” ASHRAE Transactions, Vol. 102, Part 1, pp.257_265.

D. Japikse and N.C. Blaines, 1993. Diffuser Design Technology, Concepts ETI, Inc., White River, VT.

M. R. Khorrami, F. Li and M. Choudhari, 2001. “A Novel Approach for Reducing Rotor Tip Clearance Induced Noise in Turbofan Engines” NASA Langley Research Center, American Institute of Aeronautics and Astronautics, 7th AIAA/CEAS Aeroacoustics Conference, Maastrictht, Netherlands, 28_30 May, 2001.

Nicholas C. Kernstock, “Slashing through the Noise Barrier,” Aviation Today, August 1999.

W. P. Levins, C.K. Rice, and V. D. Baxter, 1997. "Modeled and Measured Effects of Compressor Downsizing in an Existing Air Conditioner/Heat Pump in the Cooling Mode," ASHRAE Transactions, Vol. 192, Part 2, pp.22_33.

J. Proctor, Z. Katsnelson, G. Peterson and A. Edminster, Investigation of Peak Electric Load Impacts of High SEER Residential HVAC Units, Pacific Gas and Electric Company, San Francisco, CA., September, 1994.

R. G. Rajagopalan and Z. Zhang, "Performance and Flow Field of a Ducted Propeller," American Institute of Aeronautics and Astronautics, 25th Joint Propulsion Conference, AIAA_89_2673, July 1989.

G. J. Rosenquist, 1997. "Comparison of Simulated and Measured Test Data on Air_Source Heat Pumps", 3rd International Conference on Heat Pumps in Cold Climates, Wolfville, Nova Scotia, Canada, Aug. 11_12, 1997; Caneta Research, Inc., Mississauga, Ontario, Canada, November, pp.391_405.

M. Schaffer, 1991. A Practical Guide to Noise and Vibration Control for HVAC Systems, ASHRAE, Atlanta, GA.

Click here for PDF Version