Ageing of Aluminum Power Connectors Based on Current Cycle Test

The effect of current cycle on the behavior of contact resistance of clamped connectors was investigated experimentally to characterize the thermal behavior of the used clamped connector and optimizing the installation procedure in order to reduce contact resistance and ensure a lower temperature during normal operating conditions. The thermal network method is used for calculating the temperature rise of the connector when loaded by a current. The method is based on substitution of the connector geometry by a circuit consisting of thermal resistances, capacitance and heat sources. The temperature rise is determined using the network simulation program PSPICE with the corresponding thermal model libraries. The validity of the obtained results has been checked by comparing the computed values with those measured experimentally. The agreement was found satisfactory


I. INTRODUCTION
Connectors needs to fulfill reliably all technical requirements during the entire lifetime of many decades even under the assumption of continuous loading with 100% nominal current for the whole lifetime.It makes no difference if power outage was caused by complex apparatus, software or one connector failure, the consequence is the same: Extremely high costs in case of faults in industry or blackout in Grid and image loss.
The phenomena of creep and stress relaxation occurs in all metals to varying degrees.In aluminum it occurs to a higher degree than in copper, although there is some confusion as to the difference between the two terms, both creep and stress relaxation are two definitions of the same mechanism.Both describe the results of a slow atomic diffusion mechanism in the metal, which is dependent on temperature and time [1].
In the last years there has been a considerable increase in energy consumption, especially in densely populated areas.This phenomenon has caused the risk of line saturation in some areas and the consequent need to increase power lines capacity.However, it is often extremely difficult to build new distribution and transmission lines, particularly in urban areas or in regions of ecological interest, the increase of power lines capacity imposes more severe operating conditions on devices such as substation connectors, Published on January 29, 2019.G. Moustafa is with Electrical Eng.Dept.Faculty of Engineering, Jazan University, Jazan, Saudi Arabia.
involved in transmission and distribution systems, which are subjected to higher loads and have to operate at higher temperatures.[2]- [4].
Substation connectors are usually considered the weakest points in the power grid, mainly due to the poor installation practice and the lack of knowledge of their degradation rate [5].These facts often involve the difficulty to predict the useful life of a component.Moreover, it is worth mentioning that the mechanical, metallurgical, thermal and electrical processes involved in the establishment and the maintenance of the electrical contact are very complex and nowadays there is a lack of a unified model which describes the phenomena that occurs at contact interface.
Number of authors [6]- [8], have attempted to define criteria for evaluating contact performance.In fact, it has been suggested [8] that the observed changes in contact resistance relative to initial contact resistance, Rc, is a very sensitive measure of predicting performance.Intuitively one would expect that changes in resistance exceeding several times Rc would warn of substantial degradation and possible contact instability.
This paper is discussing Experimentally the ageing of aluminum connector, a current cycle test was made for several types of clamped and parallel groove connectors where the current is applied intermittent to produce a temperature around 125°C for the conductor used in the test.The performance criteria are based on measurements of the joint resistance and temperature.
Furthermore, the thermal network method based on OrCAD Capture is used for calculating the temperature rise of the connector when loaded by a current.
The experiments were carried out in Institute of Electrical Power Systems and High Voltage Engineering of the Dresden University of Technology -Germany.

A. Sample preparation
Three shapes of clamped type connectors straight, T and parallel groove connectors are shown in Fig. 1 a, b and c respectively.The connectors are made of AC-AlSi7Mgo.Straight connectors have two different shapes of the groove (serration and as cast).Two ACSR (Aluminum conductor steel reinforced) conductors (2.5 m long, 31.7 mm diameter) as shown in Fig. 2

B. Circuits diagrams and experimental procedures
Samples of clamped connectors Fig. 1 were assembled to connect aluminum conductor, with three connectors of each type.Tightening torque of values 20, 35 and 46 Nm were applied to determine the effect of increasing torque on the joint resistance.
The joint resistance of the connector samples was measured at room temperature by a micro-ohmmeter.The test points (point M and N shown in Fig. 3) are obtained by metallic straps around the conductor at a distance of 25 mm apart from two sides of the connector [9].It is a common practice that aluminum conductors should be abraded with a fine steel brush before assembled by the connector.

C. Current cycling test
Temperature cycling tests (at 125 ºC and 150 ºC) conducted on connectors that were originally rated for 70 ºC operating temperature have detected the electrical and thermal deterioration in most types of connectors [2] A test frame was designed and built of wood Fig. 4. The connectors and the test circuit are shown in Fig. 5.All tests were carried-out using ACSR conductor.The current was applied to yield a 125 o C temperature rise of the conductor, and the time required to reach this degree was recorded.The applied current is interrupted and the connector is left to cool again to room temperature.The joint temperature was continuously monitored and recorded every two minutes by a computer system with the help of two thermocouples inserted and fixed in small holes drilled into each connector side at a distance 3 mm from the clamp surface.One measurement point for each aluminum conductor in addition to two measurement points for recording room (ambient) temperature were established.The total number of temperatures measuring points is 20.Power analyzer was used to measure the temporal variation of the current during the cycle test via a current transformer.
According to the thermal network (part 4), it was found that a current of 1350 amperes is enough to yield the 125o C temperature rise aver one cycle (90 min ON-period and 90 min OFF-period) as shown in Fig. 6.The DC joint resistance was measured before starting the test and after every 50 cycles.The joint temperature was continuously monitored every 2 minutes.The test was continued for 150 cycles.

D. Joint resistance measurement
The joint resistance of this connector includes contact resistance between strand wires and sleeve, bulk resistance of the sleeve, and strand wires as well as contact resistance among individual wires of the strand wire.
Joint resistance of the above collected failed and new sample is evaluated at room temperature, the test points (point A and B shown in Fig. 3 were obtained by metallic straps around the conductor at a distance of 25mm apart from two sides of the sleeve connector According to DIN 61284 :1995.

III. EXPERIMENTAL RESULTS
The calculated performance factors K-is expressed by Eq. ( 1) as the ratio of the joint resistance divided by the resistance of an unjointed part of cable of the same length [3], [11].
Performance factor K = R_j/( R_L ) (1) A. Effect of increasing torque on the contact resistance Fig. 7 shows the relation between the performance factor and torque for different cover numbers of as cast straight clamp, namely 2, 3 and 4-cover.The curves have illustrated that the performance factor is almost independent of torque for the same cover number.The measurements show that the performance factor significantly decreases on increasing the number of covers from 2 to 3 with little decrease on increasing the number of covers from 3 to 4. This is due to a small contact area and highly current distortion in the case of 2 covers in comparison with the case of 3 and 4 covers clamp.The results of the joint resistance during current-cycling tests are summarized in Fig. 9, where the joint resistance of all samples increases with the increase of the number of current cycles.The increase in the contact resistance can be described [5] as follow When a connection was being made, the surface asperities of contacting members will penetrate the natural oxide films, thereby establishing localized metallic contact bridges.The contact pressure increased by the temperature rises resulting from the passage of electric current through the contact since aluminum has a greater coefficient of thermal expansion than stainless steel (screw) The temperature increases, and intensifies the stresses in the joint or, properly, in the metallic contact bridges.Although such increase of stress tends to make a good contact point, the stress will Tends to least resistance so that pressure increases in the axial direction of the aluminum conductor.If the magnitude of this stress exceeds the elastic limit of the aluminum conductor, the material yields and permanent deformation occurs.The effect of this deformation on the joint performance is not evident until the joint begins to cool to the initial ambient temperature, when the metallic contact bridges fail to return to their original compressed position.So, all contact resistance increase and the ratio of increase depend on contact surface area and shape, in the case of serration clamp composition another contact point is available more than that in the case of as cast clamp, this is the reason of highly increase of as cast 2 covers clamp.
The results for the variation of temperature during current cycle depict the variations of the maximum and minimum values of temperatures of 4-cover and 2-cover as cast clamps Those data were obtained by averaging the readings from two thermocouples on each clamp.The test results show that current cycling had caused no appreciable effect on the 4-cover as cast connector temperature since it remains relatively constant throughout the test The results of the current-cycling tests are summarized in Tables I and II.Table I presents the values of the joint resistance before and after current cycling test, which extended for 150 cycles.It is clear that the joint resistance increases after current cycling for all connectors and the maximum increase was found for 2-cover as cast clamp, Table II shows the temperature variation for all clamp during current cycling.significant change in joint temperature between first 50 cycle and last 50 cycle for 2cover as cast clamp connector.This behavior is attributed to the high increase in joint resistance in the case of 2-cover as cast (201.65% compared to the initial value against 10% for 4-cover clamp) moreover, the 2-cover clamp has smaller surface area to dissipate this temperature to surroundings air when compared with the 4-cover clamp.

IV. THERMAL NETWORK MODEL
A. Simulation concept.
The Thermal Network Method (TNM) is based on a substitution of an arbitrary 3D geometry by a circuit consisting of thermal resistances, capacitances and heat sources.For such a network the currents correspond to heat flow and the nodal potentials to temperatures.Due to similarity of mathematical formulations the electrical circuit programs can be used to obtain a solution.The basic advantage of the thermal network analysis is the fast computation time: steady state computations of large models can be performed within a few seconds.Therefore, the TNM is very suitable for parameter studies and become popular as a tool supporting the industrial design [10], [11] Thermal network is modeled by equivalent electrical and thermal variables Table III and calculation equation Table IV.By using the network simulation program PSPICE with the corresponding thermal model libraries and calculation equations Table IV for thermal power losses and thermal resistances for busbar the thermal networks were built Fig. 10.

B. Thermal network results
The thermal network results of power and temperature at different applied current are shown in Fig. 11 the calculation result shows a good agreement with that measured experimentally, any different between measuring and computed value for temperature may be due to instrument error or the constant room temperature 20 oC during calculation.

Fig. 3 .
Fig. 3. Schematic diagram of clamped straight connector under test for measuring joint resistance.

Fig. 7 .
Fig. 7. Performance factor as a function of Torque at different covers number of straight clamp

Fig. 1 .
Fig. 1.Thermal network model for overhead transmission line

Fig. 11 .
Fig. 11.Calculated and measured variation of temperature of overhead transmission line during on and off period of a current cycle.

TABLE I :
COMPARISON OF THE JOINT RESISTANCE OF CLAMPED CONNECTORS BEFORE AND AFTER CURRENT CYCLE

TABLE III :
RELATION BETWEEN ELECTRICAL AND THERMAL FLOW FIELD

TABLE IV
From the available data and experimental tests, The Joint resistance of clamped connector is affected by number of cover.3.Contact resistance compared with initial contactresistance showed an increase during current cycling test.4. Serrated clamped connectors showed good withstand to current cycling test has been compared with cast clamp. 5.The agreement between the theoretical and experimental temperature rise for different joint assemblies is quite good.