DEGREE PROJECT IN MECHANICAL ENGINEERING, SECOND CYCLE, 30 CREDITS
STOCKHOLM, SWEDEN 2019
Semi-active suspension seats in High speed crafts
VICTOR EKSTRÖM
KTH ROYAL INSTITUTE OF TECHNOLOGY SCHOOL OF ENGINEERING SCIENCES
1
Acknowledgements
I would like to thank my supervisor Karl Garme for continuous support during this master thesis and for yet again being my supervisor. I would thank Pahansen de Alwis for much help regarding measuresfor WBV and for general support. Lastly, I would like to dedicate this thesis to my grandfather Ingvar Johansson, who sadly passed away during this thesis. A great man that thought me everything about kindness and respect for all those around us.
2
Abstract
The working environment of a high-speed-craft (HSC) can be uncomfortable and hazardous for the crew and passengers on board. This due to the repeated exposure to high levels of vibration and shocks, that can lead to reduced performance and severe injuries to the back and neck. The most common method to reduce the vibration and shock exposure (to the crew) has been to install passive suspension seats.
The passive suspension seats have proved to reduce the vibration and shock exposure to the crew aboard HSC’s, by the measures presented in the international standards SS-ISO 2631-5 and SS-ISO 2631-1. The applicability of these measures towards HSC has been debated for a long time, especially the limit values presented in them for maximum vibration and shock exposure. Also, the limit values that are set are often quickly exceeded despite the use of passive suspension seats.
The purpose of this thesis is to investigate the possible improvement of reduction of vibrations and shocks by the use of some sort of actively controlled suspension, and to identify what measures are best suited for comparing passive suspension seats with actively controlled suspension seats.
A study is conducted where several different measures are evaluated. The measures aim at capturing vibrations and shock aboard HSC’s, the effect they have on the human body from long time exposure and immediate exposure.
A semi-active suspension system is chosen as the most suitable suspension system and is compared to a passive system by using several measures that are best suited for evaluating the risks of injuries associated with shocks and vibrations.
The semi-active suspension is simulated using Matlab Simulink, where the control method of continuous skyhook control is used for achieving the most efficient damping. Different mechanical set ups for semi-active damping is investigated in order to obtain limitations for the simulation program. The simulation program has seat base acceleration data as the input, and the seat acceleration, that migrate to the human body, as the output. The seat acceleration data of the semi-active seat is compared to recorded and simulated seat acceleration data of passive seats.
The result of the comparison is evaluated with the measures presented in ISO 2631-1, ISO 2631-5, BS 1987, and measures that are currently under development. The results show that a semi-active system is more superior than a passive system, but at the cost of a higher
travelling distance.
3
Sammanfattning
Arbetsmiljön hos ett höghastighetsfartyg (HSC) kan vara obekväm och farlig för besättningen och passagerarna ombord. Detta beror på den upprepade exponeringen för höga vibrationer och stötar som kan leda till minskad prestanda och kraftiga skador hos rygg och nacke. Den vanligaste metoden för att minska vibration och stöt exponering för besättningen har varit att installera passiva stötdämpande stolar.
De passiva stötdämpande stolarna har visat att de minskar vibrations och stötexponeringen för besättningen ombord på HSC, enligt de mätningsmetoder som presenteras i de
internationella standarderna SS-ISO 2631-5 och SS-ISO 2631-1. Användbarheten av dessa mätmetoder mot HSC har diskuterats under lång tid, särskilt de gränsvärden som presenteras för den maximala vibration och stötexponering. Gränsvärdena överskrids ofta snabbt även med användning passiva stötdämpande stolar.
Syftet med denna rapport är att undersöka en möjlig förbättrad reducering av vibrationer och stötar genom användning av någon form av aktivt kontrollerade stötdämpande stolar och bestämma vilka mätmetoder som är bäst lämpade för att jämföra passiva stötdämpande stolar med aktivt kontrollerade stötdämpande stolar.
En studie genomförs där flera olika mätningsmetoder utvärderas. Mätningsmetoderna syftar till att fånga accelerationerna ombord på HSC och effekten på människokroppen till följd av långvarig exponering och omedelbar exponering av vibrationer och stötar.
Ett semi-aktivt fjädringssystem väljs som det mest lämpliga fjädringssystemet och jämförs med ett passivt system genom att använda flera mätmetoder som är bäst lämpade för att utvärdera riskerna för skador i samband med stötar och vibrationer.
De semi-aktiva stötdämpande stolarna simuleras med Matlab Simulink där kontrollmetoden Skyhook kontroll används för att uppnå den mest effektiva dämpningen. Olika tekniker för halvaktiv dämpning undersöks för att få begränsningar för simuleringsprogrammet.
Simuleringsprogrammet har accelerationssignaler som inmatning och de reducerade
vibrationerna som utmatning, vilket jämfördes med uppmätta och simulerade data för passiva stötdämpande stolar.
Resultatet utvärderas med användning av flera mätvärden presenterade i ISO 2631-1, ISO 2631-5 och mätvärden som är under utveckling. Resultaten visar att det semi-aktiva systemet är mer effektivt än ett passivt system, men till kostnad av att stolen förflyttar sig mera under sjögång.
4
Table of content
Acknowledgements ... 1
Abstract ... 2
Sammanfattning ... 3
1 Introduction ... 6
1.1 Background ... 6
1.2 Scope ... 8
1.3 Objectives ... 8
1.4 Method ... 8
2 Shock and vibration evaluation methods ... 10
2.1 ISO 2631-1 and ISO 2631-5 ... 10
2.2 British Standard 6841:1987 BS (1987) ... 15
2.3 Impact count index (ICI) ... 19
2.4 Most Probable Extreme Acceleration Peak (MPEAP) ... 20
2.5 Method decision... 23
3 Passive suspension seats ... 25
3.1 Mechanical seat model for a passive seat... 25
4 Active suspension seats ... 29
4.1 Active suspension systems ... 29
4.2 Semi-active seat modelling ... 31
4.3 Existing semi-active suspension systems. ... 32
4.3.1 Magnetorheological dampers limitations ... 38
4.4 Semi-active system model. ... 39
5 Results ... 42
5.1 Vibration and shock measures... 42
5.2 Performance of the adaptive system ... 42
6 Discussion and Conclusion ... 50
6.1 Measure conclusion ... 50
6.2 The difference in result between the passive and the semi-active system ... 51
6.3 Further work and implementation into HSC... 53
7 References ... 55
Appendixes ... 57
Appendix A: Evaluation Data. ... 57
2009 sea-trial data, 74 kg ... 58
2009 sea-trial data, 92 kg ... 60
5
SAP, 74kg ... 62
SAP, 92kg ... 64
MAP, 74kg ... 66
MAP, 92kg ... 68
HAP, 74kg ... 70
HAP, 92kg ... 72
6
1 Introduction 1.1 Background
The working environment of high-speed-crafts (HSC) can be uncomfortable and hazardous for the crew and passengers on board. This, due to the repeated exposure to high levels of vibration and shocks that can lead to impaired performance, motion sickness, disturbed motor skills, visual impairment, increased risk of fatigue but also severe injuries to the spine and/or neck Arbetsmiljöverket (2015). The severity of these effects depends on three main factors;
the magnitude of the vibrations/shocks, the frequency of the vibrations/shocks, and the exposure time.
The problem with these vibrations and shocks can be further broken down into three stages.
The first one is that when operating HSC (regardless of sea state), there will be
vibrations/shocks in the hull. The second stage is that these vibrations/shocks will migrate into the human body in a certain way, either by a seat or simply from standing on the deck.
The third stage is that the human body will react in some way to these vibrations/shocks. In summary, shocks and vibrations will occur during operations of HSC, they will migrate into the human body and that will affect us in a negative way.
There is a potent difference between vibrations and shocks and thus also a difference in how they affect our bodies and how we evaluate them. Vibrations are a mechanical phenomenon where oscillations occur around a point of equilibrium. The oscillations can vary from being periodic to random motions that contain several frequencies. For HSC, random motions are most common. A mechanical shock is a sudden acceleration, where the acceleration or force is subject to extreme rates with respect to time Griffin (1990). For HSC this happens rather often in high speeds. The magnitude of the shocks in the hull for HSC depends on several factors; the hull shape, speed, wave height and how the cockswain operates the HSC. In a rather extreme scenario with high waves and high windspeed (significant wave height of 2 meters, 15 𝑚/𝑠 winds) and at a speed of 24 knots for a 7 meter HSC, a maximum value of 29 g in the hull was measured Ekström (2017).
In order to protect the crew from these vibrations and shocks, The Swedish coast guard, naval rescue and other Swedish naval organizations, uses shock absorbing seats that improve the ride comfort and reduce whole body vibrations and shocks during harsh driving conditions at sea. These seats are denoted as passive suspension seats which means that the suspension system have a fixed spring and damper configuration (see chapter 3 for mechanical set up).
During the spring and summer of 2017, two sea trials were conducted in order to evaluate the shock absorbing performance of these seats Ekström (2017).
During the sea trials in 2017, it was concluded that a softer/stronger damper does not necessarily lead to improved comfort for the driver and that the driver weight has a significant effect on the performance of the chock absorbing seat.
7 During the sea trials in the spring of 2017, the test was conducted by using a passive auto- adjustable damper for the driver. The sea trials were conducted with two different weight groups. The test showed that the higher weight had a positive effect on the driving comfort.
Simulations where weight, spring coefficient and damper coefficient was altered, showed that neither a higher nor lower value of the damper coefficient necessarily lead to an improve of the ride comfort Ekström (2017). In more general terms, neither a stronger nor weaker damper lead to a clear improvement of the ride comfort.
The second sea trial was conducted in the late summer of 2017 with an adjustable damper, where the driver adjusted the damper to best suit their weight. The sea trial was conducted with 3 weight groups. Again, a heavier weight performed better and for the lightest weight, of 74 kg, the damping seat showed an increase in vibrations transmitted to the body Andersson (2017).
For both sea trials it was concluded that the seats did reduce the health risks associated with shocks and vibrations by reducing the vibration dose value (VDV) and reducing root mean square (RMS) value, which are two measures commonly used for assessing risk to overall health associated with shocks and vibrations BS (1987). Despite this, the results for the seats are still alarming in terms of overall health risks.
These results are up for debate since a long time as there has been a discussion about the measures for shocks and vibrations in the marine working environment. This mainly with regards to VDV and RMS. The European Union (EU) legislation, the EU Directive 2002/44/EC, EU (2002) states maximum requirements for daily exposure of whole-body vibrations (WBV) regarding VDV and RMS with regards to the ISO standards. The requirements are expressed in terms of action values and limit values of the daily vibration exposure.
If the action value is exceeded, the employer should implement measures to reduce the vibration exposure. In the case when the limit value is exceeded, the employer should take immediate action to reduce the vibration exposure to levels below the limit value or stop the operation from continuing. This has mainly been applied in land transports. Surprisingly for sea transport, the EU directive has made an exception from the limit values. This has possibly led to a reduced focus on the working environment for the crew, but also increased insight that the limit and action value are unsuitable for the environment of HSC at sea. As seen in Garme, Burström, & Kuttenkeuler (2011) the limits set by the EU directive EU (2002) for an 8-hour workday is sometimes reached within 30 minutes for HSC crew during rough driving conditions. It has been argued that RMS underestimates the effects of WBV containing shocks. Since shocks increases the risk for spinal injury, arguably more than continuous vibrations Burström, Nilsson, & Wahlström (2011), it is essential in the assessment of WBV exposure on HSC is done with a measure that takes shocks into consideration.
8 The question that has risen from this is; can a more actively controlled system perform better than a passive system in terms of reducing the negative health aspects associated with
operating a HSC due to vibrations/shocks and how usable are the evaluation measures that are used today for assessing the health effects associated with whole body vibrations?
1.2 Scope
The scope of this thesis is to conclude how an active mitigating system would compare to a passive system and choose measures for comparing the two systems. This will be achieved by creating a simulation model for an active mitigating system and comparing the performance to recorded data from previous experiments with passive seats and simulated passive seats in terms of acceleration reduction. In order to compare the data, different forms of evaluation measures will be analysed in order to determine the best measures for comparing the results from the simulations. The limitations of this thesis are that the simulations only consider movements in the vertical direction, friction is neglected, and that the simulations do not tell us much about the how the design of the system should be built. Note, the acceleration data provided for the passive seat does not represents data from a specific manufacturer, but for passive seats that follows the same mechanical set-up as presented in chapter three.
1.3 Objectives
The questions that this thesis seeks to answer are the following:
• Which measures are most suitable for comparing the performance between different suspension systems?
• How accurate are the most commonly used measures for evaluating health risks associated with shocks and vibrations?
• What is an active damping system? Are there different types and what type is most suitable for the HSC environment?
• Is active damping a better option than passive damping in the HSC context?
1.4 Method
The method for achieving the objectives stated above will be conducted in several steps, with the end goal to have a comparison between a simulated active suspension system towards measured data and simulated data of a passive suspension system and a clear motivation for the choice of measures.
The simulation program modelling the active system is based on the program created by Katrin Olausson, Olausson (2012). The program is also used to simulate a passive system by solving the equations of motion presented in chapter 3. This is done numerically by the aid of the program MATLAB and MATLAB Simulink.
9 A research study on the different types of measures that captures vibrations and shocks is done to determine what factors are of importance when evaluating the performance of the different damping systems. This also shows which measures are most suited of the HSC environment.
A research study on different types of active suspension is done in order to understand the various types of active suspension systems, control algorithms and their different theoretical limitations.
Finally, the active system that seems most suitable for the HSC working environment is simulated and compared to the passive system. This is done by the measures decided upon in the study.
10
2 Shock and vibration evaluation methods
Ideally, in order to compare the passive suspension system with the active suspension system it is necessary to define the variables that are important to measure and that can be measured accurately for all vibration and shock exposing environments, and define limits for the daily/yearly/life exposure of vibrations and shocks. How the human body responds to vibrations and shocks is a complex phenomenon, which has no single, easily detected or predictable consequence to the human body. Research regarding human response to WBV has been conducted for decades and there are multiple methods to choose from that tries to capture the consequences of WBV with regards to measurable parameters. These parameters range from the characteristics of the persons being exposed, characteristics of the
vibrations/shocks, exposure time, direction of the vibrations and the coxswain position (standing, sitting, etcetera). For practical reasons, the main influential parameters that can be measure with good accuracy and that influences the way the human body response, are the acceleration (magnitude and frequency) and exposure time Griffin (1990).
Throughout decades of experiments, many national and international institutions have been working on standards for quantifying mechanical vibration and repeated shocks and
developing methods that provide limit values that can be used for evaluating the working environments of shocks and vibration in relation to human health and performance. Some of these institutions are International Organization for Standardization (ISO), European
Committee for Standardization (CEN), British Standards Institute (BSI), American National Standards Institute (ANSI), Deutsches Institut fur Normung (DIN) and Japanese Industrial Standards Committee (JISC).
In addition to these institutions, many universities and laboratories have been studying the effects of vibration on the human body. Many of these experiments have been conducted with the purpose of investigating human response to pure sinusoidal vibrations, but there have been few experiments regarding random signals containing several shocks Griffin (1990), which is more common in the HSC working environment. Statistical methods for the acceleration data have also been developed. For example, the Impact Count Index (ICI) Dobbins, Myers, & Dyson (2008) and extreme value analysis (Most probable largest (MPL), value of acceleration or probability of exceedance shocks) Griffin (1990).
In this chapter, several methods from these different institutions and universities will be presented and discussed in order to find a suitable set of measures for the passive and active system.
2.1 ISO 2631-1 and ISO 2631-5
The International Organization for Standardization (ISO) is a worldwide federation of national standard bodies. The main purpose is to promote worldwide industrial and commercial standards and by doing so aid in the creation of products and services that are
11 safe, reliable and of good quality ISO (1997). To this date, the most common measures for evaluating the risks and health aspects associated with shock and vibration exposure is by the measures defined in the International Standards. More specific ISO 2631-1 and ISO 2631-5.
ISO 2631-1 has been used in the EU legislation, Directive 2002/44/EC of the European Parliament for the assessment of the level of human exposure to vibration. Since Sweden is a member of the EU, the Swedish work environment authority uses the measures stated by the ISO standards Arbetsmiljöverket (2019).
ISO 2631-1
Root mean square (RMS)
Root mean square of the weighted accelerations. RMS is a basic measure for ISO 2631-1. It is defined as,
𝑅𝑀𝑆 = {1
𝑇∫ [𝑎𝑤(𝑡)
𝑇 0
]2𝑑𝑡}
1 2
(2.1)
RMS can also be calculated for an 8-hour exposure, 𝑅𝑀𝑆8ℎ = 𝑅𝑀𝑆(𝑇) ∙ [8
𝑇]
1
4 (2.2)
Where 𝑎𝑤 is the frequency weighed acceleration and 𝑇 is the duration time. In ISO 2631-1 it is stated that for RMS there is an action and limit value for an 8-hour exposure. The limit value describes the level for when the employer must take immediate action to reduce the daily exposure and the action value is when the employer must take action to reduce the daily exposure.
Table 1: Action and limit values for 𝑅𝑀𝑆8ℎ 𝑅𝑀𝑆8h
Limit value 1.15 𝑚/𝑠2
Action value 0.5 𝑚/𝑠2
Vibration dose value (VDV)
Since the RMS value poorly displays the effect of shock in an acceleration signal, the VDV has been defined for signals having a significant shock content and should be used if the crest factor, defined as the maximum instantaneous peak value of the signal divided by its RMS value, exceeds a value of 9. The VDV is commonly used within the HSC community. VDV is defined as,
𝑉𝐷𝑉 = {∫ [𝑎𝑤(𝑡)
𝑇 0
]4𝑑𝑡}
1 4
(2.3)
12 VDV for eight hours is calculated as,
𝑉𝐷𝑉8ℎ = 𝑉𝐷𝑉(𝑇) ∙ [8 𝑇]
1
4 (2.4)
Where again 𝑎𝑤 is the frequency weighed acceleration and 𝑇 is the duration time. Similar as for RMS there is action value and limit value for VDV.
Table 2: Action and limit values for 𝑉𝐷𝑉8ℎ VDV8tim
Limit value 21 𝑚/𝑠1.75
Action value 9.1 𝑚/𝑠1.75
Crest Factor
The crest factor is defined as the maximum measured value of a sequence, divided by the RMS of the sequence. The crest factor is used to investigate if the basic evaluation with RMS is enough for evaluating the vibration and shock exposure. If the crest factor is larger than 9, then methods such as the VDV measures should be applied ISO (2004).
It has been noted that for certain types of vibration that are characterized with having occasional shocks with large values, the basic evaluation method such as RMS might underestimate the health effects, even if the crest factor is smaller than 9. It is therefore recommended to use other measures if there is insecurity in the results for the RMS and there are also other recommended ratios that can be studied to determine if there is a need for another measure.
Maximum transient vibration value (MTVV)
The maximum transient vibration value is defined as the maximum read value of the running RMS at a specific time. MTVV is used as another indicator when using more than the basic evaluation methods. This is done by the following correlation.
𝑀𝑇𝑉𝑉
𝑎𝑤 = 1.5 (2.5)
If the ratio is exceeded, then alternative method to the RMS should be used.
The running RMS is a measure used when the crest factor is over 9. It takes into to account the occasional higher shocks that might appear by the use of a short integration time constant.
It defines as,
𝑎𝑤(𝑡0) = {1
𝜏∫ [𝑎𝑤(𝑡)
𝑡0 𝑡0−(𝜏)
]2𝑑𝑡}
1 2
(2.6)
13 where 𝑎𝑤(𝑡) is the instantaneous frequency weighed acceleration, 𝜏 is the integration time for running averaging, 𝑡 is the time and 𝑡0 is the time of the observation. Simplified, you could say that the running RMS is the same as RMS but for a specific time (narrow time gap).
The ISO 2631-1 provides methods for capturing the short time exposure of the WBV. There is however no direct quantitative relationship between the level of vibration and shock and the probability of injury, that is empirically proven in terms of recorded injuries connected to the limit values. It has been found that the limit values for both VDV and RMS are often exceeded Garme, Burström, & Kuttenkeuler (2011). As stated by Pahansen de Alwis, p 13, (2014), “Therefore one might think it is a matter of deciding the action and limit values.
According to the author it is a matter of deciding whether to evaluate the quality of the ride or to obtain a quantitative measure on the probability of specific injury or probability of severity based on specific type of injury or injuries”.
In conclusion, the measures presented in the ISO 2631-1 alone are not enough for evaluating WBV towards direct physical injuries. However, VDV and RMS can be used for comparing different suspension system in terms of how much of the acceleration energy is reduced from the base to the human body.
ISO 2631-5
As discussed in the introduction, one of the most common health problems associated with exposure to high accelerations is injuries to the lumbar spine. In order to enable estimation of the risk for adverse health effects due to such injuries, the ISO 2631-5 developed alternative measures to VDV and RMS. In contrast to the measures described in ISO 2631-1, which focuses more on short time exposure, the measures described in ISO 2631-5 have a long-term perspective.
This is achieved by considering the number of days and years an employee is exposed to vibration of a daily equivalent compression dose 𝑆𝑒𝑑. The 𝑆𝑒𝑑 is calculated using a lumbar spine model, which considers the ultimate strength of the lumbar spine and thus also adds a link to human endurance to the analysis ISO (2004). From the 𝑆𝑒𝑑 the risk factor 𝑅, which indicates if the risk for an injury to the lumbar spine is high or low, can be calculated as function of years of service. The daily equivalent static compression dose 𝑆𝑒𝑑 defined in ISO 2631-5 can be related to the ultimate strength of the lumbar spine.
𝑆𝑒𝑑 is calculated as follows.
𝑆𝑒𝑑,𝑧= [(𝐷𝑧,𝑑𝑚𝑧)6]
1
6 (2.7)
Where 𝑚𝑧 is a dose coefficient in the vertical direction (z) and the daily acceleration dose 𝐷𝑧,𝑑 is defined as the daily acceleration dose in the same direction (z). For the purpose of this
14 thesis, only the vibrations in the z-direction are considered, as it is the direction where the damping system act and the fact that the accelerations in the z-direction is considered larger than the other directions ISO (2004). 𝐷𝑧,𝑑 is defined as follows.
𝐷𝑧,𝑑 = [(𝑡𝑑 𝑇)
1
6(∑ 𝐴6𝑖𝑧
𝑖=1
)
1 6
] (2.8)
Where 𝑡𝑑 is the duration of the daily exposure, 𝐴 is the i:th peak of the lumbar spine response acceleration calculated from two models defined in the standard and 𝑇 is the sequence time period.
In order to describe the risks of adverse health effects in the long term, the standard defines a factor 𝑅 that, although the human variability is significant, can judge if the risk for adverse health effects on the lumbar spine is high or low. R>1.2 indicates high risks, whereas R<0.8 indicates low risks. The factor 𝑅 is defined as.
𝑅 = ∑𝑆𝑒𝑑,𝑧∙ 𝑁16 𝑆𝑢𝑖− 𝑐
𝑛
𝑖=1
(2.9)
Where 𝑁 is the number of exposure days per year and 𝑛 the number of exposure years. The constant 𝑐 represents the static pressure due to gravitational force and is suggested by the standard to be 0.25 MPa ISO (2004). 𝑆𝑢𝑖 describes the ultimate strength of the lumbar spine as function of a person’s age, 𝑏 + 𝑖, where 𝑏 is the age of when the person started to work and 𝑖 is the years of active duty that follows.
𝑆𝑢𝑖 = 6.75 − 0.066(𝑏 + 𝑖) (2.10)
Note that the R value is said to be valid if acceleration peaks do not exceed 40 𝑚/𝑠2. Due to this, the American standard International (2007) has defined that if the maximum value of 4.7 MPa for 𝑆𝑒𝑑,𝑧 is surpassed, then R is not a valid.
The R-factor methods do provide more input data that captures the human factors described in BS 1987. The method assume that the posture of the coxswain is correct during the entire sea trial, which is hard to verify and thus limits the method. However, it is a good measure for the risk of injury, especially since it can predict short term as well as long term adverse health effects on human spine which is missing in other measures.
The standard is still being developed, by incorporating more biological factors into the R- factor, such as design population and gender de Alwis (2014). Still, these methods need to be further developed to address the healing effects, disc injury and the effect of posture.
15 In summary, the R-factor provides a quantitative relationship between the vibration
environment and the risk of injury. The basis of evaluation is lumbar spine response to vibrations containing multiple shocks. It is limited due to the assumptions of the coxswain posture and current health. The method is being developed to further address the healing effects, disc injury and the effect of posture, but it still captures more factors than most methods and can be used to illustrate the time difference for different suspensions in terms of reaching an arbitrary limit value.
2.2 British Standard 6841:1987 BS (1987)
The scope of the British Standard 6841 BS (1987) is to provide a guide that gives measures for quantifying vibration and repeated shocks in relation to human health, interference with activities, discomfort, the probability of vibration perception and the incidence of motion sickness BS (1987). The guide is applicable to motions transmitted to the whole body through the supporting surfaces regardless of the surrounding environment. The British standard is very similar to the ISO standards as it uses RMS and VDV for the measures of Shocks and vibrations. Like the ISO standards, RMS may be used when the crest factor does not exceed 6.0. The crest factor is defined as the maximum acceleration peak divided by the RMS.
The vibrations are measured in z-direction, which is further explained in figure 1 below.
Figure 1: local and global coordinate system for a coxswain.
Observe the seated position, where we have local z-axis which is the direction through the body, and a global z-axis which is the z-direction perpendicular from the seat. For simplicity,
16 these two are often assumed to be the same, but realistically, this will not be the case for longer operations of HSC. Depending on the driving position of the coxswain, the effects of the vibrations will vary. This is where frequency weighted signals provide the opportunity of uniform reporting of different vibration environments depending of the direction of the vibrations through the human body.
Frequency weightings amplify the frequencies that are hazardous for human health and suppress the non-hazardous frequencies from the acceleration data. This increases the opportunity for the evaluation method to capture the accelerations of hazardous frequencies with less disturbance. Different frequency weightings are used based on the different axis of vibration. Special frequency weightings are used for evaluation of low frequency vibration affecting motion sickness. The manner which vibration affects humans is dependent on the vibration frequency content. Health, discomfort and perception are assessed using a
frequency range from 0.5 Hz to 80 Hz. Human performance assessment is only applicable to vibrations with dominant motions in the frequency range from 1.0 Hz to 80 Hz, and a
frequency range from 0.1 Hz to 0.5 Hz is used for motion sickness assessment. These frequency weightings are also applied in the measures in ISO 2631-1 and for ISO 2631-5.
In addition to the frequency weightings, the BS 6847 lift the importance of biological factors to the human response to vibration. The biological factors are denoted as instrict variables and include age, sex, size, fitness and prior injuries. The standard emphasizes the importance of these factors, but at the same time states that there is no empirical proof of injuries related to these factors with great accuracy. Meaning that it is for example clear that with higher age, the risk of injuries increases, but a limit age is very hard to set. It is also hard to collect data as it requires medical history and possibly other forms of data collection such as blood pressure and pulse.
There is no clear definition of health in the standard. The probabilities and extents of specific health effects from prolonged exposures to whole-body vibration have not been established.
The reason expressed in BS 1987 is that there is a shortage of conclusive evidence relating specific injuries to definite causes. Therefore, the standard states that there is not possible to provide a definitive dose-effect relationship between whole-body vibration and injury or health damage. Since the standard is almost 30 years old, this statement should be up for debate. There is collected subjective data concerning vibration magnitudes which cause discomfort and pain. This may give some indication of the possibility of injury for various conditions, although it is recognised that the experience may not necessarily correlate with pathological damage. So far, the subjective data has given indication that RMS values at 15 𝑚/𝑠2 or higher causes severe discomfort.
The standard also assesses the vibration effect on muscular control and vision activities. Most available literature is concerned with the specific effects of vibration on the coordinated control of hand movements and vision. Again, there is no quantification of fatigue or any consequential effects on performance which may occur during or following prolonged
vibration exposures. This might be because humans have a great ability to compensate for the
17 effects of adverse environments. According to the standard, vibration effect on human
performance is evaluated using weighted RMS values and states that “the weighted acceleration magnitude in any axis should not exceed 0.5 𝑚/𝑠2” for accurate hand manipulation and/or for precise vision.
The British standard uses RMS and VDV as the basis for measuring the vibration entering the human body. Different measures for evaluating the effects of the vibration depends of the driving position of the coxswain and what you want to evaluate such as health, performance, comfort, perception and motion sickness. Depending on the value of the crest factor, the BS 1987 standard recommends using VDV for crest factor values over 6. The preferred measure indicated in BS 1987 is VDV. Especially as the expected VDV (eVDV) is recommended for crest values below 6.
𝑒𝑉𝐷𝑉 = [(1.4 ∙ 𝑎)4 ∙ 𝑏]0.25 (2.13) Where eVDV is the estimated vibration dose value, a is the RMS value and b is the duration in seconds.
The standard states that for vibration environments containing repeated or single high shocks, which is the case for most HSC, VDV and eVDV methods provide better approach for
evaluating the severity of the vibration environment than RMS. Similar as for RMS, there is no clear connection between the dose values and injuries to health for VDV. It is doubtful how much the threshold level 15𝑚/𝑠1.75 given in the standard is applicable for HSC.
This method can be used to compare quality of two rides by the means of comparing VDV and RMS values. Its mentioned repeatedly in the standard that the methods provided can be improved further with long term experimental data and statistical methods that provide better indication of the quality of ride. The main issue is however not the lack of statistical data, it is the vague definition of the word “health”. This is because the standard does not define health with respect to any pathological aspect. As described in de Alwis, p14 (2014) “Therefore it simply gives an implication of the health effect as an injury to any part of the body during exposure to vibration. This method does not provide any quantified relationship between the vibration environment and the type of injury.”.
One unique feature of the British standard is that it investigates the human performance for
“activities”. This mainly concerns muscle control of the hands and vision and how these two activities can be reduced by WBV. The standard provides a guide for evaluating the vibration effects but provides no clear limits or dose values. This is however an important aspect as optimal performance for vision and muscular control are required onboard HSC in the
military, coast guard and search and rescue. Similarly, in the case of power boat racing, riders try to reach the maximum possible speed where no other effects come into mind rather than the eagerness of winning the race. So, hand control activities and the human vision described in the standard are very important aspects.
18 Comfort is, as mentioned in the standard, very subjective, and for the purpose of comparing acceleration data from passive seats with computer simulated active seats, comfort evaluation is not possible.
When using this standard to calculate the RMS and VDV measures, it is stated that frequency weighting is very important as the measures will focus more on certain frequencies and thus the possibility for overlooking certain values might occur. It is still debatable as to what extent it provides reasonable results for the high-speed marine craft. Further, the suggested limits for RMS is not motivated enough for evaluating the suspension systems in terms of risks of injury associated with WBV. For the purpose of comparing two different suspension systems, the limit values are not that important, as RMS and VDV can still be used for comparing how much energy is transmitted from the base of a boat to the human body in terms for each suspension system.
The British standard does not differentiate much from the ISO standards when it comes to the mathematical functions that are used for assessing WBV. A potential reason for this is that ISO 2631-1 was developed 10 years after the BS 1987 and could potentially base a lot of its measures on BS 1987. For this thesis where only a comparison is made between different damping systems, the BS 1987 does not provide much additional information about the how much better one system is compared to the other, as both VDV and RMS is presented in the ISO standards as well. The reason why the BS 1987 is important to include is that it lifts critical aspects of VDV and RMS as a standard measure. As mentioned, for the ISO standards
“The main purpose is to promote worldwide industrial and commercial standards” ISO (1997).
This is the main purpose for all standards and in order to provide such worldwide standard, you need methods that are easy to measure and provides consistent results. Acceleration peaks, frequencies and time are the three main factors used for WBV evaluation and is captured in both VDV and RMS, but as mentioned in BS 1987, does this provide equal results for all humans? The BS 1987 lifts the importance of biological factors when evaluating the injury risk of the human body regarding WBV. Age, weight, length, gender etcetera are factors that are easy to collect, but prior injuries and chronical illness is data that is hard to mass collect and create a framework for, nevertheless it is important. The input data for the two damping systems assumes identical backgrounds factors such as weight of the driver, length etcetera. If a protype were to be built, you would never find two identical test subjects that could sit next to each other in order to have the same acceleration exposure. If the VDV and RMS values for a future prototype would outperform a passive system, but the test subjects still felt more “ill” than the passive system test subject, then it might be of interest to install a camera during the test to analyse posture and examine the biological differences between the two test subjects. It is safe to say that neither BS 1987 or the ISO standards assumes identical biological background factors when RMS and VDV were developed, but the point is rather that a complement to these measures in terms of biological data would be a great addition when comparing the results of VDV and RMS towards the subjective thoughts of the coxswains.
19
2.3 Impact count index (ICI)
The purpose of the Impact Count Index (ICI) is to provide information about the severity of the shocks during operation of a HSC, by means of measured accelerations. This method is presented as an alternative method to VDV and RMS as it describes motion as repeated shocks rather than vibration. It is intended according to Dobbins to be helpful for designers, operators and owners when deciding the shock mitigation (suspension) seats for their HSC in an early stage as well as for hull designs Dobbins, Myers, & Dyson (2008).
The procedure for ICI is the following.
The accelerations are bandlimited, and no frequency weightings are applied. The vertical accelerations are recorded at the point of interest (seats, hull, stern, etcetera) and are then assigned to a ‘bin’. A bin is an interval that covers a narrow range of peak acceleration values. For example, with a bin range of 1 g: 1-1.999g, 2-2.999g, 3-3.999g, etcetera.
Preliminary work with this method has showed that a suitable range for each bin is 0.2 g (Dobbins, Myers, & Dyson, 2008). The bins are distributed over the entire range of peak acceleration values of the signal. The result is a cumulative sum of impacts – the IC – being in each bin. Dobbins, Myers, & Dyson (2008) states that early tests have showed that
depending on the sea-state, low magnitude accelerations represent the general motions of the boat, which are not considered shocks. Therefore, only the acceleration data above 1.6g are selected to present the impact count index. The results of the Cumulative sum of the impacts is plotted as the percentage of the total number of impacts compared to each bin against the impact magnitude, se figure 2 below.
Figure 2: Example plot with bins plotted against number of impacts. Image source (Dobbins, 2008, p 3)
Then the ICI criteria is defined as the percentage of impact count less than a specified percentage such as 95% (ICI95%) or 99% (ICI99%) of the total impact count. See figure 3.
20 Figure 3: Example plot with bins plotted against ICI, Image source (Dobbins, 2008, p 3)
According to Dobbins, Myers, & Dyson (2008) the ICI method is a rather effective method as it does not require any complicated calculations and only one kind of input data. In this method there are no clear limits for vibration as well as no known relationship defined between the impacts and the risk of injury. The limits for the bins are vaguely defined in Dobbins, Myers, & Dyson (2008) as well the connection for the 1.6 g limit towards the sea- state. Also, the process of dividing up the acceleration files into peak values is rather
complicated as the acceleration files are often very messy with disturbances. This method can be used for comparing quality of two rides. It states that “Further work is ongoing to link the IC/ICI analysis methodology to indices of Motion Induced Fatigue (MIF) and acute and chronic musculoskeletal injury” Dobbins, Myers, & Dyson (2008).
The method of dividing up the acceleration data into to bins for a certain range is useful for comparing the reduction of the acceleration peaks for different intervals for the passive seat and the active seat. Dividing up the acceleration data into specified interval from 0 to the maximal recorded acceleration peak and visualising its distribution with a histogram will show how the peak values differ in size between the active and passive seat and how the accelerations peaks are distributed. For example, it might show that the active seat has the majority of its peaks below 3g, compared to the passive seat having the majority of its peaks above 3g. This will not tell which acceleration peaks has increased or decreased in value, but it tells about the general distribution of all the acceleration peaks.
2.4 Most Probable Extreme Acceleration Peak (MPEAP)
This measure is developed at KTH by de Alwis & Garme (2018). The scope of the method is to indicate the severity of vibration and shock exposure onboard HSC, at a given point in time, during real time operations.
21 The background to this is that health and safety aspects for the coxswain and the crew has been in focus as well as human performance for the more extreme environments for HSC.
The task is complicated for many reasons and one of the main issues is the different time scales for the exposure-effect relationship. The adverse health effect to the human body caused by WBV is expected to come from years and months of exposure, while acute injuries come from instantaneous impacts. Psychological fatigue, reflexes and impaired work
performance is expected to decay within hours.
This measure presented in de Alwis & Garme (2018) addresses the risk of acute injuries due to instantaneous impacts. Different measures are combined in the method in order to analyse the vibrations and shocks aboard and in the method the measures are updated every second.
The method analyses 60 seconds vibration exposure history, as 60 seconds acceleration bins, every second. Then it determines the severity of expected high-intensity instantaneous impacts while separately computing the acquired severity due to the accumulated vibration exposure. The output is a severity index, intended to be displayed to the crew by an indicator that displays three colours: red, yellow and green that represents intensities high, medium and low respectively. See figure 4 below.
Figure 4: Colour indicator MPEAP. (de Alwis, p9 (2018))
The measures that are used to derive the indicator are RMS and MTVV. The input acceleration signal is frequency weighted for the z-direction. This weighting is related to health, comfort and perception in ISO 2631-1. The frequency weightings are designed including band-limiting filters of 0.4 Hz high-pass and 100 Hz low-pass.
The indicator is predictive, which means that is has the ability to foresee the magnitude of the latent acceleration peaks, based on the accumulated data from RMS and MTVV. The RMS value computed for the last 60 seconds indicates the energy transmitted to the human body and the MTVV for the last 60 seconds gives information about the impacts.
22 The high-magnitude short-duration impact characteristics causing acute injuries in the
vibration exposures are captured by a measure called Transient Factor (TF) computed for vertical accelerations measured over a period of 60 seconds, which is the ratio between MTVV and RMS, given in ISO (1997) and is expressed as
𝑎𝑤(𝑡0) = {1
𝜏∫ [𝑎𝑤(𝑡)
𝑡0 𝑡0−(𝜏)
]2𝑑𝑡}
1 2
(2.14)
𝑀𝑇𝑉𝑉 = max[𝑎𝑤(𝑡0)] (2.15)
𝑇𝑟𝑎𝑛𝑠𝑖𝑒𝑛𝑡 𝑓𝑎𝑐𝑡𝑜𝑟 =𝑀𝑇𝑉𝑉
𝑅𝑀𝑆 (2.16)
Figure 5 bellow illustrates an example TF and RMS computed over a 3-hour simulated acceleration history.
Figure 5: example TF and RMS plotted over time windows of 0.1 seconds de Alwis, p10 (2018).
RMS is a useful measure, but as seen in the figure, it remains quite stable, regardless of the varying values of TF. This is because RMS is not well suited for capturing the influence of large singular impacts. Therefore, in order incorporate these impacts, a new factor that captures the deviation or unpredictability of the TF is devised, called the Unpredictability Factor (UF) de Alwis & Garme (2018).
23 𝑈𝑛𝑝𝑟𝑒𝑑𝑖𝑐𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟 = √1
𝑁 ∑ [(𝑀𝑇𝑉𝑉 𝑅𝑀𝑆 )
∆𝑡𝑛+1
− (𝑀𝑇𝑉𝑉 𝑅𝑀𝑆 )
∆𝑡𝑛
]
𝑁−1 2
𝑛=1
(2.17)
In order to compute UF the acceleration exposure history is considered for real-time analysis, in this method 60 seconds. It is further segmented into micro windows of time ∆𝑡 and the ratio between MTVV and RMS for each micro window is computed.
The Severity Index (SI), an indicator of the exposure severity, it is computed using the RMS and UF of the 60 seconds acceleration bin formulated as
𝑆𝑒𝑣𝑒𝑟𝑖𝑡𝑦 𝑖𝑛𝑑𝑒𝑥 = 𝑅𝑀𝑆 ∙ 𝑈𝑛𝑝𝑟𝑒𝑑𝑖𝑐𝑡𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝑓𝑎𝑐𝑡𝑜𝑟 (2.18)
The Most Probable Extreme Acceleration Peak (MPEAP) is used to rank the vibration
exposure, characterized by SI, according to the severity order. The MPEAP can be interpreted as the statistically determined largest acceleration peak being the most probable to occur during an observation period. The relationship between SI and MPEAP has been determined by fitting the data into a function of the form: 𝑦 = 𝐴𝑥𝑏, where y is MPEAP, x=SI and the coefficients has been numerically determined (based on the most probable largest peak values from 27 simulated 3h histories Razola, Olausson, Garme, & Rosen (2016)) by least square fitting procedure. The results are the following:
𝑀𝑃𝐸𝐴𝑃 = 163.42∙𝑆𝐼1.3639 (2.19)
The MPEAP values are divided into different three severity zones green, yellow and red. The suggested levels in de Alwis & Garme (2018) is 0-4 g for the green zone, 4-6 g for the yellow zone and >6 g for the red zone. The main strengths for this method regardless of the
suggested limit values is that is captures the potential health risk from WBV by looking at real time data with good accuracy. The communication is straight forward as is colour-based with known colours that is easily recognized.
2.5 Method decision
It has been concluded in previous reports based on real sea trial data that the maximum allowed level for both VDV and RMS is surpassed rather quickly in high seas Ekström (2017). There has been a lot of criticism from the marine industry regarding the significance of these levels with regards to the marine environment. Still, RMS and VDV does capture long time exposure of the vibration energy and captures the three major aspects i.e.
acceleration peaks, frequency and time. Thus, they can be used to compare the performance of different suspensions systems in terms of percentage difference, but not cannot say with certainty how the suspension systems reduce the risk of “health” injuries due to WBW. This is similar for the R-factor in ISO (2004). The difference is that the R-factor considers more biological factors such as the age when the coxswain starts to work in the HSC environment
24 and exposure cycles over a year. Thus, is it will also be used and simply to compare how much longer theoretically a driver can operate with the semi-active system compared to the passive system. The MPEAP method will be used to compare the two systems in terms of how many of the semi-active systems MPEAPS are allocated to the green, yellow and red zone compared to the passive system. It is used as it is the most effective method for evaluating potential instantaneous data.
25
3 Passive suspension seats
3.1 Mechanical seat model for a passive seat
In order to compare the passive suspension system with any form of active suspension system, it is important to understand the principles of the passive suspension system and its strengths and weaknesses.
The mechanical principle of how the passive shock-absorbing seat works can be described as a 2nd degree of freedom system, see figure 6. The figure represents a shock-absorbing seat with a spring and damper. Where the spring constant (𝑘1 in N/m), the damper constant (𝑐1 in Ns/m) and the seat mass (𝑚1 in kg) are known constants. The human body is also represented in figure 6, above the red dotted line as it is also thought to have a similar setup as a mass- spring-damper system. The rationality behind this is that the human body is not a rigid mass, it has some flexibility Olausson & Garme (2014). Therefore, the upper human body is denoted as 𝑚2 (in kg) and the weight of the legs are included in 𝑚1. 𝑘2 represents the spring stiffness of the human body and 𝑐2 represents the damping stiffness of the human body. The principle of the shock-absorbing seat means that the motion energy that arises when the hull moves, is absorbed by the shock-absorbing system, which consists of a spring and a damper.
There, the kinetic energy is converted into heat energy via the shock-absorbing system. The heat energy is then absorbed by the surrounding environment.
The spring stiffness is similar to the elasticity of a material, meaning that a higher stiffness of the spring will require a larger force in order to extend it. This means that a stronger spring will lead to a lower travelling distance for a passive suspension seat and that a heavier coxswain will have a larger travelling distance compared to a lighter one for the same spring stiffness, this was concluded in Ekström (2017). The mechanical spring stores energy proportional to its compression/stretch, but the problem with having only a spring but no damper is that the stored energy will be passed back in the form of kinetic energy. This energy travel straight out into the crews’ bodies. Therefore, something that dissipates the energy is needed.
The damper works so that it always creates a counteracting force in relation to the direction of movement of the attenuated object. When the spring releases its energy, it is done by expanding the spring to its original position, which the damper counteracts. The stored energy in the spring is then converted to kinetic energy, which is absorbed by the damper and becomes heat energy which is then absorbed by the surroundings.
But the system does not absorb all energy, some energy will be passed through the seat and into the body of the driver, in the form of WBV. It is these vibrations that create the negative health effects of the driver. The weight of the seat plays an important role here as energy can be defined as the work required to move the object a certain distance. This means that a heavier object, in this case a heavier seat, requires more work (force) to move the seat body,
26 and thus the mass of the seat and weight will affect the whole-body vibrations that goes into the driver's body.
The three constants, the weight of the seat, the spring and the damper, thus affect the mitigating ability in the seat.
This leads to the mechanical set-up in figure 6 below.
Figure 6: mechanical setup of a passive suspension seat with a human.
The forces action on the system can be described using Newtons second law, giving the forces of motion in equation 3.1. Where 𝑴 is the mass matrix , 𝒙̈ is the acceleration vector, 𝑪 is the damper constant matrix, 𝒙̇ is the velocity vector, 𝑲 is the spring constant matrix, 𝒙 is the position vector and 𝑭 is the force vector.
𝑴 ∙ 𝒙̈ + 𝑪 ∙ 𝒙̇ + 𝑲 ∙ 𝒙 = 𝑭 (3.1)
𝒙̈ = [𝑥1̈
𝑥2̈ ] (3.2)
