Neena Anil

Engineering Science and Technology, an International Journal 24 - Contents lists available at ScienceDirect Engineering Science and Technology, an International Journal j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / j e s t c h Full Length Article Six Sigma based modeling of the hydraulic oil heating under low load operation Neena Anil a, Prashantha Kini c, Saurabh Gupta b,c, Hemant Darbari d, Nisheeth Joshi b, Mahdi Khosravy e,⇑ a Chief Editor & Producer, Boston United States b Department of Computer Science, Banasthali, Vidyapith Rajasthan, India c John Deere India Pvt Ltd., India d Centre for Development of Advanced Computing, Pune, Maharashtra, India e Media Integrated Communication Laboratory, Graduate School of Engineering, Osaka University, Japan a r t i c l ei n f o Article history: Received 9 February 2020 Revised 1 October 2020 Accepted 1 December 2020 Available online 6 January 2021 Keywords: Hydraulics Heat generation Radial piston pump Variable displacement pump Case drain Check valve a b s t r a c t The hydraulic system is the backbone of industrial, construction, and agricultural machines. Its high-power density and efficiency execution make it the preferred choice for high-energy applications. This manuscript presents the evidence against the statement, Whenever the system runs to its full capacity, hydraulic oil heats up more as compared to partial loads.‘‘ Our experimental study on Injection Molding Machine (IMM) hydraulic system tries to reshape the above statement by presenting the fact where we came across that lower system loading case is causing higher hydraulic oil temperatures than full load-ing.” The concern started when the machine at lower speeds results in frequent shutdowns due to increasing operating temperature high alarm!‘‘ This article presents the study to resolve the stated issue completely. For the design evolution, a systematic diagnostic approach is displayed based on the six-sigma and the associated mathematical model. Our study concludes with an approach to find the self-balancing system for the case where the temperature increase is must be due to the higher rate of heat addition than the rate of heat rejection. Presented learning can be equally extended for the other fields of interest, as care should be taken to understand the passive system losses, which can result in elevated temperatures due to lower cooling capabilities. 2020 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Introduction The hydraulic system is the backbone of high-power universal applications of industrial, construction and agriculture from small assembly processes to integrated steel, 3D printers and paper mills [1,2]. The healthy condition of hydraulic systems is highly impor-tant as its continuous monitoring by lightweight IoT edge devices has been the topic of variety of recent artificial intelligence research [3–5]. Industries use hydraulic energy as the power den-sity of hydraulics is superior than any other available form of energy. This Hydraulic system has a sophisticated control mecha-nism to perform a precision operation to improve product quality or process quality. For example, in the construction industry which has cost-effective and easy to go nature of hydraulic fluid, the hydraulic system utilizes an incompressible fluid to produce enough working force [6] for excavation. In all precision control ⇑ Corresponding author. E-mail address:-u.ac.jp (M. Khosravy). Peer review under responsibility of Karabuk University. hydraulic systems, the fluid travels through the whole hydraulic systems which comprise of precision valves which directs the sys-tem to start, stop and flow of fluid. It could deliver excellent perfor-mance due to its high precision parts in most of the operating conditions [7] wherein to maintain the precision control and relia-bility of system is designed to have resilience for temperature, pressure ripples and contamination. Ref. [8] investigates the impact on the pressure ripple in delivery pressure due to fluid iner-tia are exhaustively modeled. This resilience is achieved by proper selection and sizing of hydraulic components, as the detailed infor-mation on hydraulic component selection and design is given by Fitch et al. [9]. This research reports a practical observation of uncontrolled variables coming into effect despite fulfilling the scientific design and principles, and despite well following recommended guide-lines. It is the case while performance is with useful work diversity of working functions, the hydraulic system places demands and causes the increasing heat of the incompressible fluid. Ref. [10] typically describes the properties of hydraulic oil (incompressible fluid) to consider for performing precision applications. Despite https://doi.org/10.1016/j.jestch- -/2020 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). N. Gupta, P. Kini, S. Gupta et al.Engineering Science and Technology, an International Journal 24 - the research works on fault diagnosis of hydraulic systems [11– 13], this research is on fault reasoning. 1.1. Hydraulic universally adopted principles 1- Complicated system for conversion and transfer of energy for useful work needs the observation of oil heating to reduce the downtime of machines. Heating of hydraulic sys-tem is due to the operational inefficiencies which cause due to the loss of input power in the heat (PL) and can be expressed as [14–16], below equation has been taken to explain more on the deciding parameters affects the system efficiency by attributing the losses to different components. Ploss ¼ Ploss;pump þ Ploss;valves þ Ploss;plumbing þ Ploss;actuatorsð1Þ Distribution of the power loss in the pump, valves, plumbing, and actuators are respectively Ploss;pump þ Ploss;valves þ Ploss;plumbing and P loss;actuators and can be seen in the Fig. 1. Power loss contribution from individual subsystem is clearly visible in Fig. 1, which clearly enables the designed to plan for overall system cooling system. 2- Also, System leakage flow and associated pressure drop is one of the major factors for various component losses and need to be quantified for assessing system cooling require-ments. Eq. (1) in the context of Figs. 2 and 3 represents the losses of a typical system due to: I. The type & size of the pump (in this case a radial piston pump), II. Plumbing sizes for connection of pump & valves with reservoir, III. Type & size of the valves (in this case Electro Hydraulics), IV. Type and size of actuators (in this double acting cylinders). 3- Additionally, information on radial piston pump from Moog controls [14–16] has details on pump performance, effi-ciency, leakage, power requirements etc. These catalogs have also information on control requirements and applica-tion information. Details related to pump case flow is taken from these documents. Daniel Gronberg et al. [17] studied on prediction of case temperature of piston pumps through an analytical model. Further, Study from David Nathaniel et al. [18] has power con-sumption study on IMM and other vehicles which gives insight on power consumed during idle condition and during load. 1.2. Problem formulation For an efficient hydraulic system not to get overheat, the dissi-pation of the heat (H Dissipation) must be greater than the total input power heat loss (Total Input Power Heat Loss): Fig. 1. Distribution of hydraulic power into the useful and loss components. Fig. 2. IMM block diagram. Objective function – H Dissipation > Total Input Power Heat Loss Case 1 - H Dissipation < Total Input Power Heat Loss: The result would be increase in oil temperature very fast and eventually system shut down with degraded performance. Case 2 - H Dissipation = Total Input Power Heat Loss: the result would be increase in oil temperature slowly considering the effect of residual temperature and eventually system shut down. Although this shut down will take longer time. Mostly job rejection on machine. Case 3 - H Dissipation > Total Input Power Heat Loss: Ideal operating condition, requires investment and costly arrangements to achieve this. Trade off is required between cooling arrangements and desired quality of Job. The approach is to make a striking balance between investment required to achieve case 3 and quality of job. This eventually will improve the operating efficiency of plant. In reference to this objective function solution is proposed by Brendan Casey et al. [19]. He has presented a study on heat gener-ation in typical hydraulic system and explained to methods to resolve the same. (Comme – 1 & 2) However, the literature studies evidently show the flow of oil and associated heating is controlled by mature installed costly technologies. It varies in the machine according to their use (the relation between job quality and invest-ment), Further caution should be exercised to increase the perfor-mance level of the whole hydraulic system to avoid an undesired breakdown during the work. In References [1,7,8], two methods are generally described to control over-heating problem, first is decrease heat load where second belongs to increase heat dissipa-tion. The integrated mechanism of reservoir fluid level, airflow around the reservoir and heat exchanger ensures the above meth-ods of controlling heat problem. 1.3. Whys analysis based on literature study ‘Whys’ are an interrogative technique used iteratively for the cause-and-effect identification detailing a particular problem. The ultimate goal of this universal method is to identify the root cause of a defect or problem by repeating the question ‘‘Why?”. Each answer forms the basis of the next question. Symptom – There is overheating in hydraulic oil on low power cycle Why? Symptom – There are parasitic losses unaccounted (not consid-ered while designing universal machine) Why? Symptom – Since machine is universal and intended for varied industries and application, optimization for one application was not possible. Why? 12 N. Gupta, P. Kini, S. Gupta et al.Engineering Science and Technology, an International Journal 24 - Fig. 3. IMM schematic. Symptom- Usage of machine for intended environment was not considered while installing cooling and switchgear system for effi-cient operation and job quality. Why? Symptom – Being first of this kind of industry requirement, this is the learning evaluation. System shall take care for low power cycle losses. Why? Root Cause - In the practical condition, it turned out that low power cycle accounts for more than 50% of the operating life whereas the machine is designed for optimum efficiency on higher load cycles. 1.4. The Objective the presented research work Studies from T. Jeff Earley et al. [20] have information recom-mended oil temperature for a typical vehicle and also explain chal-lenges in maintaining the same. Based on a literate survey it is surprisingly concluded that there is not a tangible approach for ensuring the protection of the oil system for low loading or speed conditions, which may have a remarkable effect on the perfor-mance of the hydraulic machines. Improvements of the control design to be essential for the uninterrupted service of heavy equipment. In the fluid power system, more flow always results in more pressure drop, which is one of the heat generation cause [21]. Fric-tional losses create it due to conduit flow surface and intermolec-ular forces [6,17,18]. The more the pressure drop, the more is the heat addition. Effect of Above VFP, if adequate cooling arrange-ments are not made, the fluid power system temperature can rise above the recommended levels. Operating the fluid power system in the recommended temperature range always results in the extended life of the system and provides optimum efficiency. The Industrial hydraulic systems usually run within a narrow range of operating temperatures to ensure the functioning of machines in the optimum viscosity range. One of the established IMM had a peculiar issue, which was totally against hydraulic principles known during the analysis. It raises the question that ‘How can any hydraulic machine can have more heat when utilises less than when it utilizes more?’ Industrial Hydraulic machinery mostly follows a strict operating temperature range of 40 LC to 60 LC, which will ensure the ideal viscosity for all the Hydraulic Parts. The viscosity of oil also is always mentioned at 40 LC. So, efforts were put in to understand the problem systemat-ically. The effort started with the temperature mapping on the IMM and data was analyzed for understanding the trends as against the past established data. From the data analysis it was clear that pump casing temperature was steadily increasing and was going out of the recommended operating temperature range. Data also indicated that the increase is more in Pump casing tem-perature than any other parts of the system. This led to the conclu-sion that head addition is happening through the pump case and needed cooling. Pump case oil is used for the lubrication of inter-nally rotating parts and depends on the leakage through clearances of pressurized parts. The pump case is also sensitive to case pres-sure since the oil is retained inside. In this paper, we tried to explain the phenomenon of overheat-ing in case of low operating load by experimenting on the Injection Molding Machine (IMM) hydraulic system [22,23]. The approach proposed follows the self-balancing theory, where net heat addi-tion and net heat removal should be matching to result in sus-tained operating temperature for the hydraulic system needs to be at its best. This is the maximization of previously described Objective function. The methodology suggested in this article is based on a customer complaint, wherein the threshold tempera-tures were often broken down and resulting in the production stoppage even when the hydraulic system was used at lower load. The experimental results confirm the fundamental role of the self-balancing mechanism in the evaluation of heat generation for a hidden case of low operating speed. The proposed approach may be a useful tool in the arsenal of the evolution of the new control mechanism. More particularly, in Section 2, we briefly describe the system detail and architecture for the sake of analyzing and developing the solution. Section 3 shows the problem solving detail that we use to validate the contribution of the proposed approach experi-mentally. Subsequently, we show the relationship between cycle time and temperature rise before proposing the solution to extend the applicability of the proposed approach and temperature mea-surement strategy. On the control of low speed overheating frame-work, Section 4 showcase the root cause identification for the proposed solution, Section 5 presents the experimental details 13 N. Gupta, P. Kini, S. Gupta et al.Engineering Science and Technology, an International Journal 24 - where pictorial issues and implementation of the proposed method are presented. The associated results on the system with and without implementation validates the contribution of the research. Finally, Section 6 concludes the paper and proposes the future scope with generalization to other systems. 2. System description and problem details Before starting the problem definition and its solution, it would be worth explaining the hydraulic system architecture and perfor-mance specifications. It would help to understand the integration of the proposed solution to the existing architecture. We have ana-lyzed the IMM [8,16,23,24], which is a hydraulic system. Its indus-trial application is extensive and can be seen to cast the components of automotive, aerospace, avionics, computer elec-tronics, medical and dental products, instrumentation, and, so forth. Yuvin Chinniah et al. [24] has created extensive document on IMM and given detailed information on safety of operations with auxiliary equipment’s attached. Broad application in various fields such as geophysics, manufacturing, computer science moti-vated us to analyze this hydraulic system for increased efficiency. 2.1. System details IMM liquefies the plastic grains raw material and pushes into the mold cavity to get the final parts. In this machine, the duration of typical mold operation cycle is dependent on the various factors such as the size of the molded part, application requirements, com-plexity of the part, the material of the part. Typical mould flow analysis of plastic material for gear is explained in detail by Mehat, et al. [7]. Etesami et al. [6] studied mold material methods and ver-ification on mold qualities. Dang et al. [10] studied on process parameter optimization of IMM. This process helps in the reduc-tion of moulding cycle time and ensures quality parts. Madan et al. [19] studied on energy consumption of IMM, and the energy consumption of the typical part is analyzed. Wortberg et al. [1], Energy conservation in Electrohydraulic IMM is explained in detail which has subsystem level power utilization and loss prediction. Typical IMM consists of a pump unit, clamping unit, an injection unit, and ejector as principal components. The pump unit is the source of energy for IMM units, where the clamping unit is used to hold the molds splits together. The injection unit is used to make the homogeneous liquid mixture of raw plastic grain and Inject into the mold. Muller et al. [20] has studied on effective heat rejec-tion of molten plastic during the injection process. The ejector is the portion of IMM which removes the solidified final product from the mold. The typical mold is made of two mating cavities, as shown in Fig. 2, which has a volumetric replica of the final component and can be closed and opened with clamping cylinders. IMM referred here has an electronic radial piston pump [14–16] which operates the clamping unit with a proportional valve and the Injection unit with on/ off valves, Figs. 2 and 3 show a typical IMM which indicate a clamping unit which holds the mold in place and the injection unit sends the homogenized plastic mixture into the mold result-ing in the molded part which ejects through the ejection system post cooling. Casoli et al. [14] studied on piston pump flow predic-tion, cavitation impact etc. and also developed numerical code for swash plate. Generally, hydraulic IMM uses two types of clamping units the toggle and the ram type. Fig. 4 explains the system archi-tecture, which shows the flow of energy, signal, and noise. As shown Fig. 4, Control system process and feeds the signal to the electronic pump, electronic valves which in turn will execute the assigned function in efficient and systematic way. Rosato et al. [22,23,26] has created excellent handbook on IMM which contains Fig. 4. System architecture. information starting from moulding machine design, process, pro-duct, materials, marketing etc. 2.2. Problem details The central control system monitors the idle conditions of the system and takes commands from the Machine Control (operator input device). The system discussed here is a typical 150 T hydrau-lic machine consisting of a radial piston electronics variable dis-placement pump operating the clamping unit with a proportional valve and the Injection unit with on/off valves. Such machines typ-ically have cycle times in the range of 5 to 15 s based on the type of parts and their complexity. The subject IMM has a 22kw motor, 80 cc pump, and 8 kW oil cooler (water-based.). Sandeep Kumar Das et al. [27] has studied on cooler selection on IMM and helps in properly sizing of cooler based on installed power. The case drain is connected directly to the reservoir with a maximum allowed the flow of 3% of pump rated flow. Based on Fig. 4, Signal transmits from Operator control to the control system, which in turn direct Signal to the Electrical Motor & Electro-Hydraulic valves of the IMM. Energy Flow happens from the Pump to IMM Sub-functional Units resulting in IMM functional cycle. Fig. 4 also shows the heat generation, due to the energy flow, and considered as the noise for the system which happens at all the IMM subfunctions and case drains. M.A.K. ALIA et al. [28] has studied heat addition to the sys-tem by throttling and also establish rate of temperature rise for a typical system. Case drain flow keeps the pump case [17] warmer and retains at an elevated temperature based on the quantum of flow. Accordingly, to Hayat et al. [29], the Cattaneo-Christov heat model has a significant impact on thermo-migration for fluid flow. This analysis shows consideration of the variability in the thickness of surface and heat rejection properties of material. This study is helpful in getting information on the noise of the system and make the system robust. Khan et al. [30] investigated the phenomenon for Jeffrey and Casson fluid for variable thermal conductivity, homogeneous-heterogeneous reactions, variable thicked surface and Cattaneo-Christov heat flux. The indicated study shows fluid heating dependency on other uncontrolled variables for which considered to be given while designing the system application-wise. Hayat et al. (2017) [31], investigated the effect of silver as well as the copper nanoparticles composition with water on prop-erties of velocity and transfer of heat in between two stretchable 14 N. Gupta, P. Kini, S. Gupta et al.Engineering Science and Technology, an International Journal 24 - rotating disks in three dimensions with ambient thermal radiation, magnetic energy, and homogeneous-heterogeneous reactions and joule heating. With this, it is easy to understand the effect of impu-rities in the hydraulic control system, subjected to high tempera-ture and pressure cycles. Hayat et al. [32], investigated how the temperature and concentration distributions are affected by dis-tinct physical flow parameters. It helps to understand hydraulic fluid circulation in an industrial environment where the fluid is cir-culated in electrical, magnetic, and heating effect of other machines. Ahmed et al. [33], investigate that increase in tempera-ture is a function of the magnetic and squeezing parameter, this will help understanding vital parameters which require attention to be included in the design of system. 3. Problem solving methods In this section, we describe how the temperature is rising in terms of the cycle time of molding – consequently, outlines the types of measuring techniques. Here, we show the effect on the temperature for two types of molds, the PET bottles and the bottle cap, as shown in Fig. 5. PET bottles are the raw parts of blow mold-ing which is used for creating bottles. Generally, IMM operates at 70% of the speed setting in pet bot-tles molds, and 30% of the speed setting in the bottle cap. When IMM is used at 70% of the speed setting, there is expected perfor-mance, and there is no oil heating. The concern started when the new part which was selected needed a slower operation of the IMM. IMM is used for making pet bottles, as shown in Fig. 6, and it usually works with the following cycle time shown in Table 1. This table also has split up of the overall cycle time of the bottle cap into the clamping operation, injection unit operation, holding, mold open, and ejection. The overall cycle time taken by bottler cap molding is 4 Secs more than the PET bottle. Considering this, the system was going through a higher dwell time and lower hydraulic flow usage. 3.1. Signal acquisition and reproducibility analysis As the case of this research work when there is a new monitor-ing system, the repeatability and reliability of the deployed mea-surements should be verified. This requires a process to be applied to measured signals for a specific analysis. As demon-strated in Ref. [34] ANOVA analysis is used for the measuring sys-tem to understand the repeatability and reproducibility of the measuring system. Ref. [35] Explains the importance of measure-ment system analysis (MSA) for data collection. Fig. 6 shows the process of identifying the root cause, which is used for verification of the reliability and repeatability of the heat measurements. After applying the required noise cancellation preprocessing like morphological filtering [36–38], and adaptation, to assess the system capability, the measurement system analysis (MSA) in a pattern of 9 3 3 is designed wherein 9 distinct measured tem-perature values are selected with 3 temperature guns and 3 runs for each temperature values, Fig. 7 depicts the experiment plan for MSA-To know the repeatability and reproducibility (R&R) of measurement system The temperature was measured via a tem-perature gun ( 60 to 500 LC range) at following locations. Fig. 5. Pet bottle and the bottle cap. Fig. 6. Problem solving method. Table 1 IMM cycle time for PET bottle and cap. Sl. No Description Cycle PETG Bottle Bottle Cap 1 Clamping operation 1.2 3 2 Injection Unit Operation 1.8 3 3 Holding 1 2 4 Mold Open 1 1 5 Ejection 1 1 Total Cycle Time 6 10 1. Clamping/Ejector Cylinder 2. Injection Cylinder/Hydro Motor/ Injection Unit Movement Cylinder (Zone B) 3. Hydraulic Pump case (Zone C) 4. System Hydraulic Plumbing (Zone D) In Fig. 7, the 10, 5, 15, 35, 45, 50, 25, 65 and 90 are the selected measured temperature values in degree C. Whereas TG1, TG2 and TG3 are 3 temperature measurement guns. R1, R2 and R3 are 3 times reading with each gun. For example temperature 10LC is measured by 3 temperature guns 3 times, Same process is repeated for the rest of the temperature values. A similar experiment is planned for measurement for Time with 9 distinct values of time in minutes. Minitab software is used for measurement system capability analysis. As described above temperature is measured using temperature Guns at Zone A, Zone B, Zone C and Zone D after every 5 min (Fig. 8). ANOVA extracts the root cause by assessment of the variations of the multiple signals measurements of the same phenomenon and thereof verifies the repeatability and reliability of the measurements. Fig. 9 Shows overall temperature mapping and a clear trend showing the temperature increase in all the above-identified loca-tions. Measurement shown in Fig. 9, started with an ambient tem-perature of 35 LC, and all the location under observation shown a steady increase in temperature with respect to temperature. Due to the efficient water-cooling system [25], external temperatures are well within the normal range in most of the areas except the Fig. 7. ANOVA – experiment plan for MSA. 15 N. Gupta, P. Kini, S. Gupta et al. Fig. 8. Data collection process. pump casing surface, which indicated that the system had a case drain flow which was not passing through the oil cooler and is steadily heating the reservoir. 3.2. Heat calculations model and mapping Fig. 10 explains the temperature rise of only pump case which is used for producing the bottle caps; the IMM working temperature started from approximately 35LC which is ambient temperature during the start of the trials. Temperature of pump case raised steadily and reached 60LC in 80 min. It prompts the ‘‘oil tempera-ture high alarm” and supports the system temperature mapping. For molding the bottle caps, IMM works around 40LC and is consid-ered as the safest operating temperature. As the operating temper-ature reaches 60LC, the system knocks to the critical zone, which should be avoided for efficient operation of the Hydraulic System. The oil viscosity grades are defined at 40 LC which in turn will indi-cate that viscosities will change for any temperature higher or below this. Considering the sensitivity of viscosity on Hydraulic parts functioning, it would become very important to monitor the same for such systems. Ploss ¼ DTVqc ð2Þ 3600t The main reason of hydraulic power loss is the pressure drop which is proportional to the amount of oil flow/ density. Eq. (2) indicates the relationship between hydraulic power loss, tempera-ture rise, and the volume of oil. P ¼ p Q ð3Þ 600 Engineering Science and Technology, an International Journal 24 - Fig. 10. Temperature results (before solution). In the above equation, Ploss is the power loss in kW, q is the den-sity of oil in kg/dm3. Specific Heat Capacity in kJ/kgK is denoted by c, where it has value 1.67 kJ/kgK for mineral oil. Tank volume in liters is denoted by V, where, DT is the temperature increase in LC. In the denominator of the Eq. (2), variable t is the operating time which is defined in the hours. Eq. (3) calculates the Hydraulic power in relation to the flow of oil and its pressure, where P is the power in kW, p is the hydraulic pressure in kg/cm2 and Q is the flow in liters per minute (lpm). Eq. (2) measures the rise in oil temperature per hour concerning the operating conditions. The power which is required for creating the pump case drain leakage can be calculated from Eq. (3). Calcu-lations performed hereafter assumes that the net heat increase due to the return line of the machine operation is compensated by the cooling effect of the oil cooler. Table 2 shows the possible combi-nation of IMM operating Pressures & catalog values for case drain leakages. It also has theoretical temperature rise per hour in oil reservoir as per Eq. (3) (due to un-cooled oil from case drain going to the reservoir). Generally, the case drain flow is assumed to be around 3.5 lpm (3% max) which is coming out of the pump and get-ting added to the reservoir steadily. According to the Piston pump catalog, 4~6 lpm case drain flow is possible for an 80 cc radial pis-ton pump. From the Table 2, it is evident that the perceivable tempera-ture increase, DT, is possible when the case drain temperature reaches the reservoir without going through the cooler. The cool-ing effects by the reservoir area have excluded in the above calculations. Fig. 9. Temperature results of machine (before solution). 16 N. Gupta, P. Kini, S. Gupta et al. Engineering Science and Technology, an International Journal 24 - Table 2 Temperature increase per hour concerning the various operating conditions. Pressure in Case Drain in Hydraulic Power in Tank Volume in Operating Time in Specific Heat Capacity in KJ/ Density Kg/ DT in Degree Bar KW KW Liters hrs. KgK dm3 C- 4. Root cause identification vertical column. Same is mentioned in the horizontal column. Kumar et al. [42] has studied in detail on fishbone diagram which Six Sigma is a disciplined, statistical-based, data-driven helps in identifying the root cause in structured manner. With this approach, and continuous improvement methodology for eliminat- matrix structure, each critical to function is analysed over others. ing defects in a product, process, or service. Peter K Fung et al. [39] In the developed matrix, each critical to the function parameter has given detailed information on Six Sigma process and benefits of is compared with the rest of the parameter and based on relation- it. In Six Sigma, Quality Function Deployment (hereafter referred to ship 1, 3, and 9 ratings are assigned. Description of ratings is as QFD) helps to prioritize actions to improve the process or pro- defined under Table 3. Importance of each critical to function is duct to meet customers’ expectations. Chan et al. [40] has given assigned in the column ‘‘importance rating”. Where 5 and 1 repre- a detailed explanation of QFD and its application on solving com- sent high to low importance, respectively. The raw score is gener- plex problems. Erdil et al. [41] also has put similar efforts in defin- ated as a sum of multiplication of relationship and importance ing QFD methodologies and also focused on non-design related rating in each column. The relative score is the percentile of each function also. QFD tool, as shown in Table 3, is used here to identify critical function, e.g. 13% represents the share if the sum of all score the top contributing root cause. Fig. 11 shows the Pareto of the raw is considered as 100%. This helps us to prioritize the importance of score, which is used to prioritize the identified root causes. Pareto critical function. Based on the relative % importance, ranking is was indicating the excess case drain, excess system leakage, and done. high ambient temperatures as the top contributing parameters Importance rating is used to prioritize the important critical to for the identified issue. Considering there is no high leakage alarm function. Importance rank 1 shows that this critical to function is from the system and the factory was under the ambient tempera- most important for temperature rise. Explain what the reader will ture of smaller than 40LC. Moreover, excess case drain is consid- observe from the numbers showing the relation between the row ered as one of the root causes. and column variables. The intersection of each row and column gives one cell with relationship number. It shows how both critical functions correlate with delivering performance. The raw score is 4.1. Quality function deployment (QFD) procedure obtained by relationship number and importance rating. Impor- tance rank is generated by relative %. Higher the relative % higher Quality Function Deployment is a focused approach to deter- the importance rank. mine essential product attributes or qualities. These are composed of customer’s nice to have, delighter, basic needs. This requires first prioritization of the attributes and qualities of product and ser- vices, and Quality function deployment organizes them to the appropriate organizational function. Therefore, QFD is the mapping of customer-driven qualities to the responsible functions of the product and services. Paper [31] describes how the QFD can be used for strategic planning in domains such as an economic, edu- cational, and military field. It is a statistical decision-making tool that helps to drive the best decision based on available resources. Ref. [41] develops a framework using QFD to demonstrate the ver- satility of QFD. The power cannot be restricted to one domain or one function. To develop Table 3, we used Fishbone diagram as explained in, we identified critical to function 6 parameters as mentioned in Fig. 11. Contributing identified root cause prioritization. Table 3 Quality function deployment for low load temperature raise root cause based on Six Sigma. IR ICS WIT HAT ECD RIU ESL ICS- WIT- HAT- ECD- RIU- ESL- Raw Score - Relative % 13% 13% 15% 31% 6% 22% Importance Rating: Relationships: 1 = Low Importance 9 = Strong 3 = Moderate Importance 3 = Moderate 5 = High Importance 1 = Weak IR = Importance Rating 0 or Blank = No Relationship 17 N. Gupta, P. Kini, S. Gupta et al.Engineering Science and Technology, an International Journal 24 - Few of the vital control parameters have identified to correlate the performance of the system. These parameters are given as follows: 1. Excess case drain, ECD, (a) 2. Excess system leakage, ESL, (b) 3. High ambient temperature, HAT, (c) 4. Improper cooler size Will Leave this, for this physical size is fixed 5. Improper water IN temperature, WIT, (h) 6. Radiation from the injection unit, RIU, (d) 5. Developed model and proposed solution In Six Sigma, Design of Experiments (hereafter referred to as DOE) is used to find out the best possible combination of input parameters which can result in determining the best possible results in outputs. Following DOE the transfer function is devel-oped. Regression is carried out on the acquired data, which gives the value of R2 greater than 0.9. It indicates the significance of the data. The main effect and the interaction plot is generated in Figs. 12 and 13 to see the impact of each variable on the system performances resulting in the optimized performance of the sys-tem. Long W. Lam1 et al. [43] studied on establishing the three-way interaction plot to establish inter system impacts. Fig. 12. The main effect and interaction plot. Vital few control parameters were identified to get the correla-tion on the performance of the system. Table 4 shows the DOE table with selected parameters and maximum and minimum value Transfer function developed. Table 4 shows a 32 run full factorial design. +1 represents the maximum value, and 1 represents the minimum value required for the variable listed in the first row of the table. The output variables are Temperature rise CR and Total temper-ature TC Transfer functions obtained through DOE are in Eqs. (4) and (5). CR ¼ 14:2 0:00195a 0:303b 0:0739c 0:571h þ 0:000724d ð4Þ TC ¼ 124 0:0052a 1:03b 0:730c 6:15h þ 0:0116d ð5Þ Fig. 12 shows the main effect plot of each variable. The main effect occurs when the mean response changes across the levels of a factor. The main effects plots compare the relative strength of the effects across factors. For example, temperature rise remains virtually the same when you move from ECD 50 to ECD 150. We used these plots to compare the main effects. Fig. 13 shows the interaction plot between variables. Interaction between factors occurs when the change in response from the low level to the high level of one factor is not the same as the change in response at the same two levels of a second factor. That is, the effect of one factor is dependent upon a second factor. We used interactions plots to compare the relative strength of the effects across factors. Table 5 shows the correlation of the transfer function equation developed from DOE. This figure shows the essential parameters of each vari-able and the combination of variables like Effect, Coef., and P-value. Variables having P-value <0.05 are significant for system perfor-mance. R-Square is >95%. Fig. 14 explains the system block diagrams post understanding of the root cause and implementation of the system. Green colour arrow represents Energy flow into the system, subsystem and com-ponents. Blue arrow represents signal flow in the system, subsys-tem and components. Red arrow represents noise in the system, subsystem and components. Based on the heat calculations and above analysis, one can conclude that there is a need to add a pump case drain to the oil cooler. Fig. 13. Interaction plot for vital few control variables. 18 N. Gupta, P. Kini, S. Gupta et al. Table 4 DOE matrix. ECD (a) ESL (b) HAT (!) WIT (b) RIU (d- Engineering Science and Technology, an International Journal 24 - Fig. 14. System architecture updated. Fig. 15. Temperature results (after solution). 5.1. Solution Since the oil cooler inlet could experience some amount of backpressure, a direct connection of the case drain to the oil cooler was not possible. Also, according to the inherent design of the radial piston pump, it cannot have a case pressure of more than 1 Bar. Thus, there is a need to have a bypass valve arrangement, as shown in the updated system architecture in Fig. 15. A Check Valve (CV) with a cracking pressure of 0.2 bar is added in the case drain, and the oil is routed through the oil cooler. The new arrange-ment ensured a steady flow of hot oil from the pump through the oil cooler, avoiding the temperature elevation. Since the pump is sensitive to back pressure, an additional bypass check valve of 0.5 bar is added. So, any surge pressure in the pump case was sent directly to the reservoir. A pump case pressure of higher than 1 Bar can result in the pump shaft leakage. We designed a robust solu-tion which won’t hamper any of the functions of the machine. Since the maximum case drain pressure is kept under 1 bar with a proposed design change, there is a violation of the pump manu-facturer’s recommendation to avoid using the check valve in the pump case drain. Refer Fig. 16 which has plot of case temperature vs duration of machine running hours after the implementation of the proposed solutions. Post-implementation the IMM was made to run for eight hours, and the system was running in the temper-ature range of 45 to 50 LC for the identified operation, as shown in Table 5 Estimated effects and coefficients. Term Effect Coef. SE Coef. T P Constants - - a- - b- - c- - h- - d- - axb- - ax!- - axh- - axd- - bx!- - bxh- - bxd- - !xb- 7.7 0.000 !xd- - hxd- - S = 2.01836. PRESS = 260.722, R-SQ = 99.25%. R-SQ (Pred) = 96.99%, R-SQ (Adj) = 98.54%. 19 N. Gupta, P. Kini, S. Gupta et al.Engineering Science and Technology, an International Journal 24 - Fig. 16. Info-rgraph of the improvements: (a) 30% system efficiency, (b) 47% reduction in operation cost, (c) 55% oil change frequency. Fig. 17. Venn diagram – fault distribution. Fig. 15. Some of the other learnings which are gathered from the supplier catalogue [7,8] were as below: 1. A pump running with low pressure and low flow needs a pump case flushing with an external flow of 4 to 6 lpm. 2. The pump suction should not have sharp angles and screwed pipes to avoid excessive pressure drop. There is a need to use pipe bends or hoses for the connection. 3. There is a need to refer the pump manufacturer catalogue for the maximum possible case drainpipe/ hose length as it directly impacts the pump case pressure when there is a higher leakage. 4. The fluid temperature in the tank must not exceed the temper-ature of the pump by more than + 25. With the above results, the rate of heat addition from the sys-tem operation and rate cooling will ensure operating temperature within the recommended temperature range (self-balancing). 5.2. Improvements The observable clear improvements according to our extensive experiments are as follows (i) system efficiency, (ii) oil changing frequency (times/year), and (iii) operating cost. The improvement for each of the above-mentioned cases is respectively by 30%, 47%, and 55%. The system efficiency is calculated based on direct and indirect overheads required for cooling of hydraulic oil. At the same time oil changing frequency reduction by 55% is due to low operating temperature which results in slow degradation of oil. Indeed, temperature plays a vital role in degradation of oil. Just in the case of oil and efficiency improvement there is an expecta- tion of 30% improvement in operating cost. Since machine is able to operate in safe temperature zone therefore contamination also reduced, and filter life is expected to improve too. Fig. 17 shows the Below Venn diagram illustrating the distribution of faults due to filter chocking, Pump degradation and oil degradation. Approx-imately 50% of the failure happens because of oil degradation and Filter chocking. Operation under recommended temperature zone helps to reduce oil degradation time, eventually filter life increases. 6. Conclusion It is uncommon to see any hydraulic system heating up at lower loads. In this report, we have discussed a rare case of IMM, wherein the system had heated up at lower loads. Here, the system leakage flow was responsible for the higher system operating tempera-tures. Through this study, the authors stress the importance of rul-ing out the possibility of a system leakage, whenever the system heats up, though it’s a rare possibility. The authors also convey that this learning applies to all hydraulic systems and is not limited to IMM. It is crucial to maintaining the right recommended tempera-ture range throughout the life of the hydraulic machines. The method explained here will be useful for both Industrial and mobile hydraulics applications. Method explained in this paper helps to identify unintended losses in the Hydraulic system like Pump casing temperature, Neutral condition losses, losses due to wear and tear, etc. Pressure transducer can be made part of the reservoir to monitor the temperature and give an alarm when the temperature is outside recommended range. Typically for Industrial machines its recommended to have temperature from 40 LC to 60 LC. Oil needs to be heated through a heater or oil pre-warming method to attain minimum operating temperature range and needs to be cooled with a cooler to keep within maximum operating range. It ensures the optimum performance and life of the system. For typical mobile hydraulic applications, recom-mended temperatures will be in the range of 20 LC to 100 LC which mostly uses oil with high Viscosity Index. There is a need to follow the user manual judiciously to avoid underperformance and breakdown of hydraulic systems. As our future research scope, we are intended to examine the hydraulic system of high power construction and agricultural machines to make the presented phenomena as universal as design evolution of hydraulic machines for efficient operation ranging from low load to high load. Main-taining a recommended temperature in a tractor can avoid prema-ture failures of Implement Pumps, Steering Valve, Brake Valve, Control Valves etc. Similarly, all the Agricultural and Construction machinery, if operated within the recommended temperature will provide extended life of parts and minimum downtime. Maintain-ing temperature in a Hydraulic press will ensure reduced 20 N. Gupta, P. Kini, S. Gupta et al. downtime due to improved reliability, optimum viscosity for effi-cient functions. Declaration of Competing Interest The authors declare that they have no known competing finan-cial interests or personal relationships that could have appeared to influence the work reported in this paper. References [1] J. Wortberg, T. Kamps, Hydraulics and all-electric injection moulding machines in practice, Plast. Europe 7 -. [2] Ravi Ramya, Chandrasekharan Rajendran, Hans Ziegler, Sanjay Mohapatra, K. Ganesh, CLSP: real life applications and motivation to study lot sizing problems in process industries, in: Capacitated Lot Sizing Problems in Process Industries, Springer, Cham, 2019, pp. 33-45. [3] N. Gupta, M. Khosravy, N. Patel, N. Dey, S. Gupta, H. Darbari, R.G. Crespo, Economic data analytic AI technique on IoT edge devices for health monitoring of agriculture machines, Appl. Intell. -. [4] N. Gupta, M. Khosravy, S. Gupta, N. Dey, R.G. Crespo, Lightweight artificial intelligence technology for health diagnosis of agriculture vehicles: parallel evolving artificial neural networks by genetic algorithm, Int. J. Parallel Prog. -. [5] N. Gupta, S. Gupta, M. Khosravy, N. Dey, N. Joshi, R.G. Crespo, N. Patel, Economic IoT strategy: the future technology for health monitoring and diagnostic of agriculture vehicles, J. Intell. Manuf. -. [6] F. Etesami, C. Mullens, R. Sahli, T. Webb, in: Proceedings of 34th International Confer, 2019, pp. 438–443. [7] N.M. Mehat, S. Kamaruddin, A.R. Othman, Modeling and analysis of injection moulding process parameters for plastic gear industry application, ISRN Industr. Eng. 2013 -, https://doi.org/10.1155/2013/869736. [8] P. Casoli, A. Vacca, G. Franzoni, G.L. Berta, Modelling of fluid properties in hydraulic positive displacement machines, Simul. Model. Pract. Theory 14 (8) -, https://doi.org/10.1016/j.simpat-. [9] Ernest C. Fitch, I.T. Hong, Hydraulic Component Design and Selection, BarDyne, 2004. [10] X.-P. Dang, General frameworks for optimization of plastic injection molding process parameters, Simul. Model. Pract. Theory 41 -, https://doi. org/10.1016/j.simpat-. [11] S. Gupta, M. Khosravy, N. Gupta, B.N. Tiwari, B. Senzio-Savino, F. Asharif, M.R. Asharif, Tractor oil pump fault diagnosis by pseudo-spectrum analysis of vehicle sound records, in: Proceedings of the 31st International Technical Conference on Circuits/Systems, Computers and Communications. [12] S. Gupta, M. Khosravy, N. Gupta, H. Darbari, N. Patel, Hydraulic system onboard monitoring and fault diagnostic in agricultural machine, Braz. Arch. Biol. Technol. 62 (2019), https://doi.org/10.1590/-. [13] S. Gupta, M. Khosravy, N. Gupta, H. Darbari, In-field failure assessment of tractor hydraulic system operation via pseudospectrum of acoustic measurements, Turk. J. Elec. Eng. Comp. Sci. 27 (4) -, https://doi.org/10.3906/elk-. [14] Radial Piston Pump Modular design for superior performance quiet and robust, RevJ, Feb 2017 http://www.moog.com/literature/ICD/MoogPumps-RKP-Catalog-en.pdf [15] Dynamic displacement and pressure control for demanding applications, February 2017. http://www.moog.com/content/dam/moog/literature/ICD/ Moog-PumpsRKP-D-Catalog-en.pdf. [16] Radial Piston Pump RKP Version, January 2010. http://www.moog.com/ literature/ICD/rkpIIpumps-om.pdf CA-. [17] D. Grönberg, Prediction of Case Temperature of Axial Piston Pumps, 2011. [18] D.N. Kordonowy, A Power Assessment of Machining Tools (Doctoral dissertation), Massachusetts Institute of Technology, 2002. [19] B. Casey, Symptoms of Common Hydraulic Problems and Their Root Causes, Machinery Lubrication, 2003. Engineering Science and Technology, an International Journal 24 - [20] Keep Your Hydraulic System Running, This article is reprinted from American Sweeper magazine, Volume 7 Number 2, 1999. [https:// www.worldsweeper.com/PreventiveMaintenance/v7n2hydraulic.html]. [21] https://www.crossco.com/blog/3-common-causes-hydraulic-overheating. [22] D.V. Rosato, M.G. Rosato, Injection Molding Handbook, Springer Science Business Media, 2012. [23] M.G. Rosato, D.V. Rosato (Eds.), Plastics Design Handbook, Springer Science Business Media, 2013. [24] Plastic Injection Moulding Machines with Auxiliary Equipment by Yuvin Chinniah, Sabrina Jocelyn, Barthlemy Aucourt, Ral Bourbonnire; IRSST Communications and Knowledge Transfer Division, MAY 2017. http://www. irsst.qc.ca/media/documents/PubIRSST/R-970.pdf. [25] Fritz Muller, Injection mold apparatus with improved heating and cooling system, U.S. Patent 5,055,025, issued October 8, 1991. [26] D.V.R. Dominick, V. Rosato, Marlene G. Rosato, Injection Molding Handbook, Kluver Academic Publishers, Norwell, Massachusetts, USA, 2000. [27] S.K. Das, K. Bhojak, Oil cooler selection for hydraulic system of plastic injection moulding machine: a review, Int. J. Sci. Res. 5 -. [28] M.A.K. Alia, T. Younes, H. Sarhan, Hydraulic Domestic Heating by Throttling, Published Online June 2010 (http://www.SciRP.org/journal/eng). [29] T. Hayat, M. Ijaz Khan, M. Farooq, A. Alsaedi, M. Waqas, Tabassam Yasmeen, Impact of Cattaneo–Christov heat flux model in flow of variable thermal conductivity fluid over a variable thicked surface, Int. J. Heat Mass Transf. 99 -, https://doi.org/10.1016/j.ijheatmasstransfer-. [30] Muhammad Khan, Muhammed Waqas, Tasawar Hayat, Ahmed Alsaedi, Colloidal study of Casson fluid with homogeneous-heterogeneous reactions, J. Colloid Interface Sci. 498 (2017), https://doi.org/10.1016/j.jcis-. [31] Tasawar Hayat, Sumaira Qayyum, Muhammad Ijaz Khan, Ahmed Alsaedi, Current progresses about probable error and statistical declaration for radiative two phase flow using Ag H2O and Cu H2O nanomaterials, Int. J. Hydrogen Energy 42 (49) -, https://doi.org/10.1016/j. ijhydene-. [32] Salman Ahmad, M. Khan, Tasawar Hayat, Muhammad Khan, Ahmed Alsaedi, Entropy generation optimization and unsteady squeezing flow of viscous fluid with five different shapes of nanoparticles, Colloids Surf., A 554 (2018), https://doi.org/10.1016/j.colsurfa-. [33] Tasawar Hayat, Arsalan Aziz, Taseer Muhammad, Ahmed Alsaedi, Active and passive controls of 3D nanofluid flow by a convectively heated nonlinear stretching surface, Phys. Scr. 94 (8) -, https://doi.org/10.1088/-/ab1307. [34] A. Yadav, Measurement systems analysis and a study of ANOVA method. [35] Assessing Measurement System Variation. Available: https://www. minitab.com/uploadedFiles/ Documents/sample- materials/ FuelInjectorNozzles-EN.pdf [36] M.H. Sedaaghi, R. Daj, M. Khosravi, Mediated morphological filters, in: Proceedings 2001 International Conference on Image Processing (Cat. No. 01CH37205), IEEE, Vol. 3, 2001, October, pp. 692–695. [37] M. Khosravi, M.H. Sedaaghi, Impulsive noise suppression of electrocardiogram signals with mediated morphological filters, in: The 11th Iranian Conference on Biomedical Engineering, Tehran, Iran, 2004, February, pp. 207–212. [38] M. Khosravy, N. Gupta, N. Marina, I.K. Sethi, M.R. Asharif, Morphological filters: an inspiration from natural geometrical erosion and dilation, in: Nature-Inspired Computing and Optimization, Springer, Cham, 2017, pp. 349–379. [39] Design for Six Sigma (DFSS) Introduction 24 January 2015, Peter K. Fung, Hong Kong http://www.hksq.org/DFSS-STP-150124-Peter.pdf. [40] Lai-Kow Chan, Ming-Lu Wu, Quality function deployment: a literature review, Eur. J. Oper. Res. 143 (3) -, https://doi.org/10.1016/S- -. [41] Nadiye Ozlem Erdil, Omid M. Arani, Quality function deployment: more than a design tool, IJQSS 11 (2) -, https://doi.org/10.1108/IJQSS-. [42] M. Pradeep Kumar, M. Krishna, Dr. N.V.S. Raju, Dr. M.V. Satish Kumar, Failure criticality analysis using Fishikawa diagram (a case study of dumpers at OCP, Ramagundam). [43] A typology of three-way interaction models: Applications and suggestions for Asian management research by Long W. Lam1 Aichia Chuang & Chi-Sum Wong & Julie N. Y. Zhu June.2018 , Springer Science + Business Media, LLC, part of Springer Nature, 2018. 21